# Query User Balance
Source: https://docs.vidgo.ai/api-manual/account-management/user-balance

api-manual/account-management/user-balance.json GET /api/user/balance
Query the credit balance of your user account

Query the current credit balance of your user account in real-time.

## Use Cases

> [card] Dashboard Display
>
> Show real-time account balance in your application dashboard

> [card] Balance Alerts
>
> Set up notifications when credits fall below a threshold

> [card] Recharge Reminders
>
> Prompt users to top up credits before running out

> [card] Usage Tracking
>
> Monitor credit consumption and track spending patterns

## Important Notes

> [!NOTE]
>
> **Real-time Balance**: The balance returned reflects your current credit amount at the time of the request.

> [!TIP]
>
> **Rate Limiting**: Avoid excessive polling. Cache the balance and refresh only when necessary (e.g., after completing a generation task).

> [!WARNING]
>
> **Credit Deduction**: Credits are deducted only when generation tasks complete successfully. Failed tasks do not consume credits.


# Error Codes
Source: https://docs.vidgo.ai/api-manual/error-codes

API error codes and response formats

This page describes common response formats for generation and native Chat Series endpoints. Use the HTTP status and the endpoint-specific error body.

## Response Format

### Success Response

Generation submission returns a response with `code: 200`. Native Chat Series responses use their protocol-specific format:

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "task-unified-1757165031-uyujaw3d",
    "status": "running",
    "created_time": "2025-11-12T10:30:00"
  }
}
```

### Error Response

Generation endpoints return the HTTP error status with a detail field. They do not wrap errors in code and error:

```json theme={null}
{
  "detail": "task_id is required"
}
```

### Native Chat Series Errors

Chat Completions and Responses use an error object; Claude Messages adds type=error; Gemini uses error.code, error.message, and error.status. Authentication failures may still use detail. Upstream error status codes and bodies are forwarded unchanged. Successful JSON responses and SSE events are not wrapped in code and data.

### Task Failed Response

When a task fails during processing, the response includes `status: "failed"` with an error message:

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "failed",
    "files": [],
    "created_time": "2025-11-25T08:50:13",
    "error_message": "The prompt violates our content policy"
  }
}
```

## HTTP Status Codes

### 200 - Success

Request completed successfully. For generation endpoints, check `data.status` to determine task state. Native Chat Series returns its JSON response or SSE stream directly.

### 400 - Bad Request

The request contains invalid parameters or data.

**Example:**

```json theme={null}
{
  "detail": "task_id is required"
}
```

### 401 - Unauthorized

Missing or invalid credentials. Authentication errors use a detail field.

### 402 - Payment Required

Native Chat Series endpoints use HTTP 402 for insufficient credits. Generation submission uses HTTP 400 for insufficient credits.

**Example:**

```json theme={null}
{
  "error": {"message": "Insufficient credits", "type": "invalid_request_error", "code": null}
}
```

### 403 - Forbidden

Access denied due to permission restrictions.

**Example:**

```json theme={null}
{
  "detail": "You do not have permission to access this resource"
}
```

### 404 - Not Found

The requested resource does not exist.

**Example:**

```json theme={null}
{
  "detail": "Task not found"
}
```

### 408 - Request Timeout

The request timed out before a response was returned. The timeout middleware returns plain text:

**Example:**

```text theme={null}
Request timeout
```

### 422 - Unprocessable Entity

Malformed JSON or a request body that is not an object may be rejected before endpoint handling. The detail field is an array of validation errors.

### 429 - Too Many Requests

Rate limit exceeded.

**Example:**

```json theme={null}
{
  "detail": "Rate limit exceeded. Please try again later."
}
```

### 500 - Internal Server Error

An unexpected server error occurred.

**Example:**

```json theme={null}
{
  "detail": "An internal error occurred. Please try again later."
}
```

### 502 - Bad Gateway

Upstream service error.

**Example:**

```json theme={null}
{
  "detail": "Upstream service unavailable"
}
```

### 503 - Service Unavailable

The requested model is temporarily unavailable. Contact support to check model availability before retrying. An unknown or disabled public model normally returns HTTP 404.

## Error Handling Best Practices

> [!TIP]
>
> **Retry transient failures carefully**: For temporary 500 or 502 errors, use exponential backoff. Retrying submission can create another task. For HTTP 503, contact support to check model availability before retrying.

> [!TIP]
>
> **Handle rate limits**: When receiving 429 errors, wait before retrying. Consider implementing request queuing.

> [!TIP]
>
> **Validate inputs**: Check your request parameters before sending to avoid 400 errors.

> [!TIP]
>
> **Monitor credits**: Generation submission reports insufficient credits as HTTP 400; native Chat Series reports HTTP 402.


# Upload File from Base64
Source: https://docs.vidgo.ai/api-manual/file-series/upload-base64

api-manual/file-series/upload-base64.json POST /api/common/upload/base64
Upload files to Vidgo API using Base64-encoded data or Data URL format

## Usage Guide

* This endpoint accepts Base64-encoded file data for upload to Vidgo API storage
* Supports both Data URL format (with MIME type prefix) and pure Base64 strings
* Ideal for uploading files that are already Base64-encoded in your application, such as canvas drawings or client-side image manipulations
* Files are immediately accessible via the returned URL and remain available for 72 hours

## Parameter Details

* **Base64 Data Format**:

  * **Data URL format**: `data:image/png;base64,iVBORw0KGgo...` (recommended)

  * **Pure Base64 string**: `iVBORw0KGgo...` (without MIME type prefix)

  * The Base64 string must not contain any whitespace, newline characters, or formatting

  * Supported file formats: **JPEG, PNG, GIF, WebP** only

  * Maximum upload limit: **1 image per request**

* **Storage Configuration**:

  * All files are automatically stored with a `temp/` prefix in the storage path

  * If you specify `upload_path: "profile-images"`, the actual path will be `temp/profile-images`

  * Files will expire and be automatically deleted **72 hours** after upload

* **File Naming**:

  * If `file_name` is not provided, the system generates a unique name in the format: `{timestamp}_{random}_{extension}`

  * Example auto-generated name: `20251229130857_a8B9cD2e.png`

## Usage Notes

* When to use Base64 upload:

  * Client-side generated images (canvas, screenshots)

  * Images already encoded in Base64 format in your application

  * Small to medium-sized images where encoding overhead is acceptable

* When NOT to use Base64 upload:

  * Large image files (Base64 encoding increases size by \~33%)

  * Files stored on disk (use stream upload instead)

  * Files available via URL (use URL upload instead)

* Ensure your Base64 data is properly formatted without any line breaks or spaces

* For persistent storage needs, download and save the file locally before the 72-hour expiration

## Rate Limits and Quotas

* **Rate Limit**: 5 requests per minute per API key

* When the rate limit is exceeded, you will receive a `429 Too Many Requests` error

* Implement exponential backoff retry logic for handling rate limit errors

## Common Error Scenarios

* **Invalid Base64 Encoding**: The provided Base64 string cannot be decoded

* **Invalid Data URL Format**: The Data URL format is malformed (e.g., missing `data:` prefix or `;base64,` separator)

* **Unsupported File Type**: The decoded file is not in a supported image format (JPEG, PNG, GIF, WebP)

* **Invalid File Data**: The decoded data is empty or corrupted

* **Authentication Error**: Missing or invalid API key in the Authorization header


# Upload File Stream
Source: https://docs.vidgo.ai/api-manual/file-series/upload-stream

api-manual/file-series/upload-stream.json POST /api/common/upload/stream
Upload files to Vidgo API using multipart/form-data format for direct file uploads

## Usage Guide

* This endpoint enables direct file uploads using the standard `multipart/form-data` format
* Supports uploading files directly from local storage with automatic file type identification
* Best suited for local file uploads from desktop applications, mobile apps, or web forms
* Files are immediately accessible via the returned URL and remain available for 72 hours

## Parameter Details

* **File Upload Requirement**:

  * The `file` field must contain binary file data

  * Supported file formats: **JPEG, PNG, GIF, WebP** only

  * Maximum upload limit: **1 image per request**

  * The system automatically identifies the file type from the binary data

* **Parameter Naming Convention**:

  * This endpoint supports both **snake\_case** and **camelCase** naming conventions

  * `upload_path` or `uploadPath` - both are accepted

  * `file_name` or `fileName` - both are accepted

  * Use whichever convention matches your application's coding style

* **Storage Configuration**:

  * All files are automatically stored with a `temp/` prefix in the storage path

  * If you specify `upload_path: "photos"`, the actual path will be `temp/photos`

  * Files will expire and be automatically deleted **72 hours** after upload

* **File Naming**:

  * If `file_name` is not provided, the system generates a unique name in the format: `{timestamp}_{random}_{extension}`

  * Example auto-generated name: `20251229130857_a8B9cD2e.png`

## Usage Notes

* When to use file stream upload:

  * Direct file uploads from user's local storage

  * Form-based file uploads from web applications

  * Mobile app file uploads

  * Server-to-server file transfers

* This is the most efficient upload method for files already stored on disk

* The multipart/form-data format is standard across all programming languages and HTTP clients

* For persistent storage needs, download and save the file locally before the 72-hour expiration

## Rate Limits and Quotas

* **Rate Limit**: 5 requests per minute per API key

* When the rate limit is exceeded, you will receive a `429 Too Many Requests` error

* Implement exponential backoff retry logic for handling rate limit errors

## Common Error Scenarios

* **Missing File**: No file was provided in the request body (422 error)

* **Unsupported File Type**: The uploaded file is not in a supported image format (JPEG, PNG, GIF, WebP)

* **Empty File**: The uploaded file has zero bytes or is corrupted

* **Authentication Error**: Missing or invalid API key in the Authorization header

## Example Usage

### cURL Example

```bash theme={null}
curl -X POST "https://api.vidgo.ai/api/common/upload/stream" \
  -H "Authorization: Bearer VIDGO_API_KEY" \
  -F "file=@/path/to/image.png" \
  -F "file_name=my-image.png"
```

### Python Example

```python theme={null}
import requests

url = "https://api.vidgo.ai/api/common/upload/stream"
headers = {"Authorization": "Bearer VIDGO_API_KEY"}

files = {"file": open("/path/to/image.png", "rb")}
data = {"file_name": "my-image.png"}

response = requests.post(url, headers=headers, files=files, data=data)
print(response.json())
```

### JavaScript Example (Node.js)

```javascript theme={null}
const FormData = require('form-data');
const fs = require('fs');
const axios = require('axios');

const form = new FormData();
form.append('file', fs.createReadStream('/path/to/image.png'));
form.append('file_name', 'my-image.png');

axios.post('https://api.vidgo.ai/api/common/upload/stream', form, {
  headers: {
    'Authorization': 'Bearer VIDGO_API_KEY',
    ...form.getHeaders()
  }
})
.then(response => console.log(response.data))
.catch(error => console.error(error));
```


# Upload File from URL
Source: https://docs.vidgo.ai/api-manual/file-series/upload-url

api-manual/file-series/upload-url.json POST /api/common/upload/url
Upload files to Vidgo API by providing a remote URL

## Usage Guide

* This endpoint enables uploading files to Vidgo API by providing a remote URL
* The system automatically downloads the file from the specified URL and stores it in Vidgo API storage
* Ideal for migrating files from external servers or integrating with third-party file sources
* Files are returned with a direct access URL and remain available for 72 hours

## Parameter Details

* **File URL Requirement**:

  * The `file_url` must be a publicly accessible URL using HTTP or HTTPS protocol

  * The remote server must allow the file to be downloaded without authentication

  * Supported file formats: **JPEG, PNG, GIF, WebP** only

  * Maximum upload limit: **1 image per request**

* **Storage Configuration**:

  * All files are automatically stored with a `temp/` prefix in the storage path

  * If you specify `upload_path: "avatars"`, the actual path will be `temp/avatars`

  * Files will expire and be automatically deleted **72 hours** after upload

* **File Naming**:

  * If `file_name` is not provided, the system generates a unique name in the format: `{timestamp}_{random}_{extension}`

  * Example auto-generated name: `20251229130857_a8B9cD2e.png`

## Usage Notes

* Ensure your source URL is publicly accessible and does not require authentication or special headers

* The download process may take a few seconds depending on the file size and network conditions

* For persistent storage needs, download and save the file locally before the 72-hour expiration

* The `file_url` and `download_url` in the response are identical and both provide direct access to the uploaded file

## Rate Limits and Quotas

* **Rate Limit**: 5 requests per minute per API key

* When the rate limit is exceeded, you will receive a `429 Too Many Requests` error

* Implement exponential backoff retry logic for handling rate limit errors

## Common Error Scenarios

* **Invalid URL**: The provided URL is malformed or unreachable

* **Download Failure**: Network timeout or connection issues when downloading from the remote URL

* **Unsupported File Type**: The file at the URL is not in a supported image format (JPEG, PNG, GIF, WebP)

* **Authentication Error**: Missing or invalid API key in the Authorization header


# Wan 2.7 Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/alibaba-wan-2-7-image-edit

api-manual/image-series/alibaba-wan-2-7-image-edit.json POST /api/generate/submit
Rework images with text instructions and up to four visual references, exploring new compositions and variations with Wan 2.7.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Rework images with text instructions and up to four visual references, exploring new compositions and variations with Wan 2.7.

## Models

* **alibaba/wan-2.7/image-edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-4 images in `image_urls`.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Default: `1024x1024`.
* Presets: `512x512`, `1024x1024`, `768x1024`, `1024x768`, `576x1024`, `1024x576`.
* For custom dimensions, supply an object with integer `width` and `height`; both must be positive.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `4`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-4 items. Each item must be an HTTP(S) URL. Order matters: refer to them as image 1, image 2, image 3, image 4 in the prompt. A single string or Base64 image is not accepted.

### Seed

* `seed`: Optional. Set the generation seed. Integer.

## Billing Factors

Billing depends on the selected model variant and output image count (`n`). Output size does not add a charge. Reference images do not add a surcharge.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.7 Pro Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/alibaba-wan-2-7-image-edit-pro

api-manual/image-series/alibaba-wan-2-7-image-edit-pro.json POST /api/generate/submit
Guide Wan 2.7 Pro edits with text and multiple reference images to develop new looks for existing visual concepts.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Guide Wan 2.7 Pro edits with text and multiple reference images to develop new looks for existing visual concepts.

## Models

* **alibaba/wan-2.7/image-edit-pro** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-4 images in `image_urls`.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Default: `1024x1024`.
* Presets: `512x512`, `1024x1024`, `768x1024`, `1024x768`, `576x1024`, `1024x576`.
* For custom dimensions, supply an object with integer `width` and `height`; both must be positive.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `4`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-4 items. Each item must be an HTTP(S) URL. Order matters: refer to them as image 1, image 2, image 3, image 4 in the prompt. A single string or Base64 image is not accepted.

### Seed

* `seed`: Optional. Set the generation seed. Integer.

## Billing Factors

Billing depends on the selected model variant and output image count (`n`). Output size does not add a charge. Reference images do not add a surcharge.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.7 Image Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/alibaba-wan-2-7-text-to-image

api-manual/image-series/alibaba-wan-2-7-text-to-image.json POST /api/generate/submit
Turn scene and design ideas into images with Wan 2.7, with custom canvas sizes and up to four variations per prompt.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn scene and design ideas into images with Wan 2.7, with custom canvas sizes and up to four variations per prompt.

## Models

* **alibaba/wan-2.7/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Control the number of output images with `n`.

Use the corresponding image-edit endpoint for reference images.

## Output Options

### Image Size

* `size`: Optional. Default: `1024x1024`.
* Presets: `512x512`, `1024x1024`, `768x1024`, `1024x768`, `576x1024`, `1024x576`.
* For custom dimensions, supply an object with integer `width` and `height`; both must be positive.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `4`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Seed

* `seed`: Optional. Set the generation seed. Integer.

## Billing Factors

Billing depends on the selected model variant and output image count (`n`). Output size does not add a charge.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.7 Pro Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/alibaba-wan-2-7-text-to-image-pro

api-manual/image-series/alibaba-wan-2-7-text-to-image-pro.json POST /api/generate/submit
Create image variations with Wan 2.7 Pro, using detailed prompts to guide the subject, style, and composition.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create image variations with Wan 2.7 Pro, using detailed prompts to guide the subject, style, and composition.

## Models

* **alibaba/wan-2.7/text-to-image-pro** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Control the number of output images with `n`.

Use the corresponding image-edit endpoint for reference images.

## Output Options

### Image Size

* `size`: Optional. Default: `1024x1024`.
* Presets: `512x512`, `1024x1024`, `768x1024`, `1024x768`, `576x1024`, `1024x576`.
* For custom dimensions, supply an object with integer `width` and `height`; both must be positive.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `4`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Seed

* `seed`: Optional. Set the generation seed. Integer.

## Billing Factors

Billing depends on the selected model variant and output image count (`n`). Output size does not add a charge.

Credits are deducted on submission and refunded if the task fails.


# Z-Image Turbo
Source: https://docs.vidgo.ai/api-manual/image-series/alibaba-z-image-turbo

api-manual/image-series/alibaba-z-image-turbo.json POST /api/generate/submit
Text-to-image generation and single-image editing with Z-Image Turbo

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

`alibaba/z-image/turbo` generates images from text and edits a single reference image.

## Input parameters

* `prompt`: Required string containing 1 to 1000 characters. Describe the image to generate or the edits to apply.
* `image_urls`: Optional array containing exactly one HTTP(S) image URL. Providing this field selects image editing.
* `size`: Required for text-to-image and optional for image editing. Values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`.
* `enable_safety_checker`: Optional boolean. Defaults to `true`.

## Text to image

```json theme={null}
{
  "model": "alibaba/z-image/turbo",
  "input": {
    "prompt": "A ceramic vase on a wooden table in soft daylight.",
    "size": "1:1",
    "enable_safety_checker": true
  }
}
```

## Image editing

```json theme={null}
{
  "model": "alibaba/z-image/turbo",
  "input": {
    "prompt": "Add soft studio lighting to this product photo.",
    "image_urls": ["https://example.com/reference.png"],
    "enable_safety_checker": true
  }
}
```


# FLUX.2 Flex Edit
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-2-flex-edit

api-manual/image-series/blackforestlabs-flux-2-flex-edit.json POST /api/generate/submit
Use FLUX.2 Flex to rework images or combine references, guiding the subject, style, and composition through text.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Use FLUX.2 Flex to rework images or combine references, guiding the subject, style, and composition through text.

## Models

* **blackforestlabs/flux-2-flex/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-8 images in `image_urls`.
* Choose from `1K`, `2K` output resolutions.

## Output Options

### Image Size

* `size`: Required. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `auto`. `auto` uses the aspect ratio of the first image in `image_urls`.

### Resolution

* `resolution`: Required. Accepted values: `1K`, `2K`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send the input fields listed on this page. Each input URL is a public HTTP(S) URL.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **3-5,000 characters**.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-8 items. Each item must be an HTTP(S) URL.

## Verified Example

This example was generated through the Vidgo API and is also the default example in the playground. Source and result images are hosted on the Vidgo CDN.

![Reference image 1](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-flex/edit/v1/01/input-01.png)

![Reference image 2](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-flex/edit/v1/01/input-02.png)

![Reference image 3](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-flex/edit/v1/01/input-03.png)

![Pip and the Glass Garden](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-flex/edit/v1/01/output.jpg)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "E4MB5ERGH360DFJI",
    "status": "finished",
    "files": [
      {
        "file_type": "image",
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-flex/edit/v1/01/output.jpg"
      }
    ],
    "created_time": "2026-09-26T11:07:59",
    "error_message": null,
    "progress": 100
  }
}
```


# FLUX.2 Flex Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-2-flex-text-to-image

api-manual/image-series/blackforestlabs-flux-2-flex-text-to-image.json POST /api/generate/submit
Explore image concepts with FLUX.2 Flex, using descriptive prompts to shape composition and style in 1K or 2K output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Explore image concepts with FLUX.2 Flex, using descriptive prompts to shape composition and style in 1K or 2K output.

## Models

* **blackforestlabs/flux-2-flex/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Choose from `1K`, `2K` output resolutions.

## Output Options

### Image Size

* `size`: Required. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `auto`.

### Resolution

* `resolution`: Required. Accepted values: `1K`, `2K`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send the input fields listed on this page.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **3-5,000 characters**.

## Verified Example

This example was generated through the Vidgo API and is also the default example in the playground. Source and result images are hosted on the Vidgo CDN.

![Tide Pool Field Notebook](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-flex/text-to-image/v1/02/output.jpg)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "SBR3R4D3GOFPXM1Z",
    "status": "finished",
    "files": [
      {
        "file_type": "image",
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-flex/text-to-image/v1/02/output.jpg"
      }
    ],
    "created_time": "2026-09-26T11:07:50",
    "error_message": null,
    "progress": 100
  }
}
```


# FLUX.2 Pro Edit
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-2-pro-edit

api-manual/image-series/blackforestlabs-flux-2-pro-edit.json POST /api/generate/submit
Combine visual references with text instructions in FLUX.2 Pro to edit scenes, transfer styles, and retain recognizable subjects.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Combine visual references with text instructions in FLUX.2 Pro to edit scenes, transfer styles, and retain recognizable subjects.

## Models

* **blackforestlabs/flux-2-pro/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-8 images in `image_urls`.
* Choose from `1K`, `2K` output resolutions.

## Output Options

### Image Size

* `size`: Required. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `auto`. `auto` uses the aspect ratio of the first image in `image_urls`.

### Resolution

* `resolution`: Required. Accepted values: `1K`, `2K`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send the input fields listed on this page. Each input URL is a public HTTP(S) URL.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **3-5,000 characters**.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-8 items. Each item must be an HTTP(S) URL.

## Verified Example

This example was generated through the Vidgo API and is also the default example in the playground. Source and result images are hosted on the Vidgo CDN.

![Reference image 1](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-pro/edit/v1/01/input-01.png)

![Reference image 2](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-pro/edit/v1/01/input-02.png)

![Reference image 3](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-pro/edit/v1/01/input-03.png)

![Dumpling Lesson](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-pro/edit/v1/01/output.jpg)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "G4SI8DYZSIKS9UV5",
    "status": "finished",
    "files": [
      {
        "file_type": "image",
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-pro/edit/v1/01/output.jpg"
      }
    ],
    "created_time": "2026-09-26T10:37:28",
    "error_message": null,
    "progress": 100
  }
}
```


# FLUX.2 Pro Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-2-pro-text-to-image

api-manual/image-series/blackforestlabs-flux-2-pro-text-to-image.json POST /api/generate/submit
Create detailed images with FLUX.2 Pro, using natural language to direct subjects, composition, and visible text at up to 2K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create detailed images with FLUX.2 Pro, using natural language to direct subjects, composition, and visible text at up to 2K.

## Models

* **blackforestlabs/flux-2-pro/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Choose from `1K`, `2K` output resolutions.

## Output Options

### Image Size

* `size`: Required. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `auto`.

### Resolution

* `resolution`: Required. Accepted values: `1K`, `2K`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send the input fields listed on this page.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **3-5,000 characters**.

## Verified Example

This example was generated through the Vidgo API and is also the default example in the playground. Source and result images are hosted on the Vidgo CDN.

![Night Ferry in the Rain](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-pro/text-to-image/v1/01/output.jpg)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "FUI8N7NW2JLV34RB",
    "status": "finished",
    "files": [
      {
        "file_type": "image",
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-2-pro/text-to-image/v1/01/output.jpg"
      }
    ],
    "created_time": "2026-09-26T10:30:10",
    "error_message": null,
    "progress": 100
  }
}
```


# FLUX 3 Klein Edit
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-3-klein-edit

api-manual/image-series/blackforestlabs-flux-3-klein-edit.json POST /api/generate/submit
The FLUX 3 Klein Edit API edits or combines 1–10 input images from a natural-language instruction and returns 1–4 results at 1K, 2K or 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

FLUX 3 Klein Edit changes or combines existing images from a plain-language instruction. Send one image to restyle, retouch or replace elements, or up to ten to compose a product into a new scene or merge several references into one picture, with output at 1K, 2K or 4K.

FLUX 3 Klein is the compact member of Black Forest Labs' FLUX 3 image family, built for fast generation and instruction-based editing. Vidgo exposes it as two model IDs: [Text to Image](/api-manual/image-series/blackforestlabs-flux-3-klein-text-to-image) and [Edit](/api-manual/image-series/blackforestlabs-flux-3-klein-edit).

## Models

* **blackforestlabs/flux-3-klein/edit** — Image edit with FLUX 3 Klein, 1–10 input images, 1–4 results.

## Features

* Image edit from a natural-language instruction.
* Multi-image composition with up to 10 inputs, referenced by order in the prompt.
* Three resolution tiers: `1K`, `2K`, `4K`; presets or a custom size.
* PNG, JPEG or WebP output, 1–4 images per request.

## Required parameters

* `prompt` (string): The edit instruction. Refer to inputs by position, for example "the first image" and "the second image". Must not be blank.
* `image_urls` (string\[]): 1–10 images. Public HTTP(S) URLs, or base64 data URIs that Vidgo uploads first.

## Optional parameters

* `n` (integer): Number of images, `1`–`4`. Default `1`.
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`. See [Resolution](#resolution).
* `size` (string): Preset or custom `WIDTHxHEIGHT`. Default `1:1`. See [Resolution](#resolution).
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.
* `enable_safety_checker` (boolean): Upstream safety checker. Default `true`.

## Resolution

* With a preset `size` (`1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`), the image is scaled so its long edge matches the tier: `1K` → 1024 px, `2K` → 2048 px, `4K` → 4096 px.
* With a custom `size` such as `1920x1080`, each side must be 256–4096 and a multiple of 16, and you must send `resolution` matching the long edge: up to 1024 → `1K`, up to 2048 → `2K`, above 2048 → `4K`. A mismatch returns 400.

| `resolution` | Long edge (presets) | Custom long edge |
| - | - | - |
| `1K` | 1024 px | ≤ 1024 px |
| `2K` | 2048 px | 1025–2048 px |
| `4K` | 4096 px | 2049–4096 px |

## Output

The finished task returns one file per requested image (`n`) in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Edit one image**: turn a daytime street into a rainy night scene, `16:9`, `1K`.
* **Combine two images at 2K with callback**: put the product from the first image into the scene from the second, `n: 2`, WebP, with a `callback_url`.

## Notes

* Results with many references can vary in pose and viewpoint consistency; describe the subject clearly and keep the number of inputs as low as the task allows.
* An empty `image_urls` array or more than ten images returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and `n`. Text-to-image and edit use the same rates. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates. The number of input images does not change the price.


# FLUX 3 Klein Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-3-klein-text-to-image

api-manual/image-series/blackforestlabs-flux-3-klein-text-to-image.json POST /api/generate/submit
The FLUX 3 Klein Text to Image API generates 1–4 images from a prompt at 1K, 2K or 4K, with 7 aspect ratios or custom sizes and PNG, JPEG or WebP output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

FLUX 3 Klein Text to Image creates images from a prompt at 1K, 2K or 4K, using seven aspect-ratio presets or an exact custom size. It suits product shots, marketing visuals, concept art and high-volume image pipelines where fast turnaround matters.

FLUX 3 Klein is the compact member of Black Forest Labs' FLUX 3 image family, built for fast generation and instruction-based editing. Vidgo exposes it as two model IDs: [Text to Image](/api-manual/image-series/blackforestlabs-flux-3-klein-text-to-image) and [Edit](/api-manual/image-series/blackforestlabs-flux-3-klein-edit).

## Models

* **blackforestlabs/flux-3-klein/text-to-image** — Text-to-image with FLUX 3 Klein, 1–4 images at 1K, 2K or 4K.

## Features

* Text-to-image from a single prompt.
* Three resolution tiers: `1K`, `2K`, `4K`.
* Seven aspect-ratio presets, or a custom `WIDTHxHEIGHT` from 256 to 4096 px per side.
* PNG, JPEG or WebP output, 1–4 images per request.

## Required parameters

* `prompt` (string): The image you want. Must not be blank.

## Optional parameters

* `n` (integer): Number of images, `1`–`4`. Default `1`.
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`. See [Resolution](#resolution).
* `size` (string): Preset or custom `WIDTHxHEIGHT`. Default `1:1`. See [Resolution](#resolution).
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.
* `enable_safety_checker` (boolean): Upstream safety checker. Default `true`.

## Resolution

* With a preset `size` (`1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`), the image is scaled so its long edge matches the tier: `1K` → 1024 px, `2K` → 2048 px, `4K` → 4096 px.
* With a custom `size` such as `1920x1080`, each side must be 256–4096 and a multiple of 16, and you must send `resolution` matching the long edge: up to 1024 → `1K`, up to 2048 → `2K`, above 2048 → `4K`. A mismatch returns 400.

| `resolution` | Long edge (presets) | Custom long edge |
| - | - | - |
| `1K` | 1024 px | ≤ 1024 px |
| `2K` | 2048 px | 1025–2048 px |
| `4K` | 4096 px | 2049–4096 px |

## Output

The finished task returns one file per requested image (`n`) in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **1K square image**: a product photo at `1:1`, `1K`, PNG.
* **Custom 4K size, two JPEGs, with callback**: `size: 3840x2160` with `resolution: 4K` (long edge above 2048), `n: 2`, JPEG, with a `callback_url`.

## Notes

* This ID takes text only; `image_urls` returns 400. To edit or combine images, use [FLUX 3 Klein Edit](/api-manual/image-series/blackforestlabs-flux-3-klein-edit).
* Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and `n`. Text-to-image and edit use the same rates. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Flux Dev Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-dev

api-manual/image-series/blackforestlabs-flux-dev.json POST /api/generate/submit
The Flux Dev Text to Image API generates detailed PNG or JPEG images from a text prompt, with preset aspect ratios, 1:1 HD or a custom WIDTHxHEIGHT canvas.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Submit a Task

Send `model: "blackforestlabs/flux/dev"` and an `input` object to `POST /api/generate/submit`. Authenticate with `Authorization: Bearer VIDGO_API_KEY`.

* `prompt`: Required string. Describe the image or editing instructions using at least one character after trimming.
* `size`: Output canvas, default `1:1`. Choose `1:1`, `1:1 HD`, `4:3`, `3:4`, `16:9`, `9:16`, or a custom `WIDTHxHEIGHT` such as `1024x1536`. Custom width and height are positive integers.
* `n`: Positive integer specifying the output image count, default `1`.
* `output_format`: `png` or `jpeg`, default `png`.

## Retrieve Images

The submission response returns `data.task_id`. Use [Query Task Status](/api-manual/task-management/status) to follow the task through `not_started`, `running`, `finished`, or `failed`.

Read image URLs from `data.files[].file_url` when the status is `finished`. For `failed`, read `data.error_message`.

You can provide a top-level `callback_url` to receive the terminal result through a [webhook](/api-manual/task-management/webhooks).

## Verified Example

This example was generated through the Vidgo test API. The source and result images are hosted on the Vidgo CDN.

![The Violin Restorer](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux/dev/v1/01/output.png)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "ZOBDNFN25SVGKU85",
    "status": "finished",
    "files": [
      {
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux/dev/v1/01/output.png",
        "file_type": "image"
      }
    ],
    "created_time": "2026-09-25T11:20:26",
    "error_message": null,
    "progress": 100
  }
}
```


# Flux Dev Image to Image
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-dev-image-to-image

api-manual/image-series/blackforestlabs-flux-dev-image-to-image.json POST /api/generate/submit
Edit one source image using a text prompt with PNG or JPEG output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Submit a Task

Send `model: "blackforestlabs/flux/dev/image-to-image"` and an `input` object to `POST /api/generate/submit`. Authenticate with `Authorization: Bearer VIDGO_API_KEY`.

* `prompt`: Required string. Describe the image or editing instructions using at least one character after trimming.
* `image_urls`: Required array containing one public HTTP(S) source-image URL. The source image guides the edited output.
* `n`: Positive integer specifying the output image count, default `1`.
* `output_format`: `png` or `jpeg`, default `png`.

### Canvas Parameter

`size` defines the text-to-image canvas and defaults to `1:1`. Image-to-image generation uses the source image to guide the output. The request example supplies the source image, prompt, image count, and output format.

## Retrieve Images

The submission response returns `data.task_id`. Use [Query Task Status](/api-manual/task-management/status) to follow the task through `not_started`, `running`, `finished`, or `failed`.

Read image URLs from `data.files[].file_url` when the status is `finished`. For `failed`, read `data.error_message`.

You can provide a top-level `callback_url` to receive the terminal result through a [webhook](/api-manual/task-management/webhooks).

## Verified Example

This example was generated through the Vidgo test API. The source and result images are hosted on the Vidgo CDN.

![Source image](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux/dev/image-to-image/v1/01/input.png)

![A Warmer Reading Room](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux/dev/image-to-image/v1/01/output.png)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "I0RYI5JVOJC6CR8I",
    "status": "finished",
    "files": [
      {
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux/dev/image-to-image/v1/01/output.png",
        "file_type": "image"
      }
    ],
    "created_time": "2026-09-25T11:22:15",
    "error_message": null,
    "progress": 100
  }
}
```


# Flux Kontext Max Edit
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-kontext-max

api-manual/image-series/blackforestlabs-flux-kontext-max.json POST /api/generate/submit
Apply contextual image edits with FLUX Kontext Max, refining subjects and scenes while maintaining visual continuity.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Apply contextual image edits with FLUX Kontext Max, refining subjects and scenes while maintaining visual continuity.

## Models

* **blackforestlabs/flux-kontext-max** - Edit one source image with a text prompt.

## Features

* Edit one source image with a text prompt.
* Provide the source image as a single item in `image_urls`.

## Output Options

### Image Size

* `size`: Optional. Output aspect ratio. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `21:9`, `9:21`. Default: `1:1`.

### Output Format

* `output_format`: Optional. Output image file format. Accepted values: `png`, `jpg`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Input image URLs use HTTP(S) and point to publicly accessible image files.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-2,000 characters**.
* Include at least one non-whitespace character.

### Image URLs

* `image_urls`: Required. Provide exactly 1 publicly accessible HTTP(S) image URL as the source image.

## Verified Example

This example was generated through the Vidgo API and is also the default example in the playground. Source and result images are hosted on the Vidgo CDN.

![Source image 1](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-max/v1/01/input-01.png)

![Same Cellist on a Frozen Lake](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-max/v1/01/output.png)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "QBNW6JCNI530EKJV",
    "status": "finished",
    "files": [
      {
        "file_type": "image",
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-max/v1/01/output.png"
      }
    ],
    "created_time": "2026-09-26T11:02:47",
    "error_message": null,
    "progress": 100
  }
}
```


# Flux Kontext Max Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-kontext-max-text-to-image

api-manual/image-series/blackforestlabs-flux-kontext-max-text-to-image.json POST /api/generate/submit
Create images with FLUX Kontext Max, using detailed prompts to direct the subject, visual style, and framing.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create images with FLUX Kontext Max, using detailed prompts to direct the subject, visual style, and framing.

## Models

* **blackforestlabs/flux-kontext-max/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.

## Output Options

### Image Size

* `size`: Optional. Output aspect ratio. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `21:9`, `9:21`. Default: `1:1`.

### Output Format

* `output_format`: Optional. Output image file format. Accepted values: `png`, `jpg`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-2,000 characters**.
* Include at least one non-whitespace character.

## Verified Example

This example was generated through the Vidgo API and is also the default example in the playground. Source and result images are hosted on the Vidgo CDN.

![Through the Porthole of a Deep-Sea Submersible](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-max/text-to-image/v1/03/output.png)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "ENZJ08EV0WERQ3QY",
    "status": "finished",
    "files": [
      {
        "file_type": "image",
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-max/text-to-image/v1/03/output.png"
      }
    ],
    "created_time": "2026-09-26T11:05:02",
    "error_message": null,
    "progress": 100
  }
}
```


# Flux Kontext Pro Edit
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-kontext-pro

api-manual/image-series/blackforestlabs-flux-kontext-pro.json POST /api/generate/submit
Edit images with FLUX Kontext Pro using text instructions, with support for character consistency and changes to specific visual details.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Edit images with FLUX Kontext Pro using text instructions, with support for character consistency and changes to specific visual details.

## Models

* **blackforestlabs/flux-kontext-pro** - Edit one source image with a text prompt.

## Features

* Edit one source image with a text prompt.
* Provide the source image as a single item in `image_urls`.

## Output Options

### Image Size

* `size`: Optional. Output aspect ratio. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `21:9`, `9:21`. Default: `1:1`.

### Output Format

* `output_format`: Optional. Output image file format. Accepted values: `png`, `jpg`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Input image URLs use HTTP(S) and point to publicly accessible image files.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-2,000 characters**.
* Include at least one non-whitespace character.

### Image URLs

* `image_urls`: Required. Provide exactly 1 publicly accessible HTTP(S) image URL as the source image.

## Verified Example

This example was generated through the Vidgo API and is also the default example in the playground. Source and result images are hosted on the Vidgo CDN.

![Source image 1](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-pro/v1/02/input-01.png)

![Renaming a Fishing Boat in Chiloé](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-pro/v1/02/output.jpg)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "KH8OWRITIOASWYGA",
    "status": "finished",
    "files": [
      {
        "file_type": "image",
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-pro/v1/02/output.jpg"
      }
    ],
    "created_time": "2026-09-26T11:02:37",
    "error_message": null,
    "progress": 100
  }
}
```


# Flux Kontext Pro Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-kontext-pro-text-to-image

api-manual/image-series/blackforestlabs-flux-kontext-pro-text-to-image.json POST /api/generate/submit
Turn natural language descriptions into images with FLUX Kontext Pro, from illustrations to composed scenes and design concepts.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn natural language descriptions into images with FLUX Kontext Pro, from illustrations to composed scenes and design concepts.

## Models

* **blackforestlabs/flux-kontext-pro/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.

## Output Options

### Image Size

* `size`: Optional. Output aspect ratio. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `21:9`, `9:21`. Default: `1:1`.

### Output Format

* `output_format`: Optional. Output image file format. Accepted values: `png`, `jpg`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-2,000 characters**.
* Include at least one non-whitespace character.

## Verified Example

This example was generated through the Vidgo API and is also the default example in the playground. Source and result images are hosted on the Vidgo CDN.

![Storm Chasers on the Kansas Plains](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-pro/text-to-image/v1/01/output.png)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "2M5RH08SQ7GQGKRE",
    "status": "finished",
    "files": [
      {
        "file_type": "image",
        "file_url": "https://cdn.vidgo.ai/apis/models/blackforestlabs/flux-kontext-pro/text-to-image/v1/01/output.png"
      }
    ],
    "created_time": "2026-09-26T10:59:34",
    "error_message": null,
    "progress": 100
  }
}
```


# Flux Schnell Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/blackforestlabs-flux-schnell

api-manual/image-series/blackforestlabs-flux-schnell.json POST /api/generate/submit
The Flux Schnell Text to Image API quickly turns a prompt into PNG or JPEG images, with preset aspect ratios, 1:1 HD or custom WIDTHxHEIGHT sizes and batch output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Submit a Task

Send `model: "blackforestlabs/flux/schnell"` and an `input` object to `POST /api/generate/submit`. Authenticate with `Authorization: Bearer VIDGO_API_KEY`.

* `prompt`: Required string. Describe the image using at least one character after trimming.
* `size`: Output canvas, default `1:1`. Choose `1:1`, `1:1 HD`, `4:3`, `3:4`, `16:9`, `9:16`, or a custom `WIDTHxHEIGHT` such as `1024x1536`. Custom width and height are positive integers.
* `n`: Positive integer specifying the output image count, default `1`.
* `output_format`: `png` or `jpeg`, default `png`.

## Retrieve Images

The submission response returns `data.task_id`. Use [Query Task Status](/api-manual/task-management/status) to follow the task through `not_started`, `running`, `finished`, or `failed`.

Read image URLs from `data.files[].file_url` when the status is `finished`. For `failed`, read `data.error_message`.

You can provide a top-level `callback_url` to receive the terminal result through a [webhook](/api-manual/task-management/webhooks).

## Verified Example

First Light at the Cliffside Pool. This text-to-image example uses a prompt only.

```json theme={null}
{
  "model": "blackforestlabs/flux/schnell",
  "input": {
    "prompt": "An architectural photograph of an empty public seawater swimming pool built into a coastal cliff at sunrise. From a low viewpoint beside the pool, pale turquoise water fills the foreground, broad curved stone steps descend naturally into the shallow end on the right, and a simple white limewashed wall catches moving water reflections. Warm low morning sunlight reveals the rough limestone coping and delicate ripples. Beyond the far pool edge, the calm open sea and a clean level horizon remain clearly visible. Restrained Mediterranean colors, believable pool geometry, natural light, crisp tactile surfaces, peaceful everyday public architecture. One coherent full-frame photograph, no people, hotel styling, advertising, signs, lettering, logos or watermark.",
    "size": "4:3",
    "n": 1,
    "output_format": "png"
  }
}
```

![First Light at the Cliffside Pool](https://cdn.vidgo.ai/apis/models/blackforestlabs/flux/schnell/text-to-image/v1/01/output.png)


# Seedream 4.5 Edit
Source: https://docs.vidgo.ai/api-manual/image-series/bytedance-seedream-v4-5-edit

api-manual/image-series/bytedance-seedream-v4-5-edit.json POST /api/generate/submit
Edit images with Seedream 4.5 using a text prompt and 1-10 reference images, generating 1-15 edited images per request.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Edit images with Seedream 4.5 using a text prompt and 1-10 reference images, generating 1-15 edited images per request.

## Models

* **bytedance/seedream/v4.5/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-10 images in `image_urls`.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Default: `1:1`.
* Presets: `2K`, `4K`, `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `21:9`.
* For custom dimensions, supply an object with integer `width` and `height`; both must be positive.
* Custom dimensions also accept a `WIDTHxHEIGHT` string, using `x` or `X` between the positive dimensions.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `15`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Input image URLs use HTTP(S) and point to publicly accessible JPEG or PNG files.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-3,000 characters**.
* Include at least one non-whitespace character.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-10 items. Each item must be an HTTP(S) URL pointing to a JPEG or PNG image.

### Safety Checker

* `enable_safety_checker`: Optional. Set `true` to enable the content safety check or `false` to disable it. Default: `true`.

## Verified Example

This example was generated through the Vidgo test API with the `basic` request shown in the API playground. It combines three references in order: the child, the toy sailboat, and the lakeshore. The source and result images are hosted on the Vidgo CDN.

![Reference 1: child](https://cdn.vidgo.ai/apis/models/bytedance/seedream/v4.5/edit/v1/01/input-01.png)

![Reference 2: toy sailboat](https://cdn.vidgo.ai/apis/models/bytedance/seedream/v4.5/edit/v1/01/input-02.png)

![Reference 3: lakeshore](https://cdn.vidgo.ai/apis/models/bytedance/seedream/v4.5/edit/v1/01/input-03.png)

![Launching the Toy Boat](https://cdn.vidgo.ai/apis/models/bytedance/seedream/v4.5/edit/v1/01/output.jpg)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "GAPP1TYEWHR5CQVS",
    "status": "finished",
    "files": [
      {
        "file_url": "https://cdn.vidgo.ai/apis/models/bytedance/seedream/v4.5/edit/v1/01/output.jpg",
        "file_type": "image"
      }
    ],
    "created_time": "2026-09-25T16:16:53",
    "error_message": null,
    "progress": 100
  }
}
```


# Seedream 4.5 Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/bytedance-seedream-v4-5-text-to-image

api-manual/image-series/bytedance-seedream-v4-5-text-to-image.json POST /api/generate/submit
Develop visual concepts with Seedream 4.5, generating multiple images from a prompt to explore different styles and compositions.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Develop visual concepts with Seedream 4.5, generating multiple images from a prompt to explore different styles and compositions.

## Models

* **bytedance/seedream/v4.5/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Default: `1:1`.
* Presets: `2K`, `4K`, `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `21:9`.
* For custom dimensions, supply an object with integer `width` and `height`; both must be positive.
* Custom dimensions also accept a `WIDTHxHEIGHT` string, using `x` or `X` between the positive dimensions.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `15`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-3,000 characters**.
* Include at least one non-whitespace character.

### Safety Checker

* `enable_safety_checker`: Optional. Set `true` to enable the content safety check or `false` to disable it. Default: `true`.

## Verified Example

This example was generated through the Vidgo test API with the `basic` request shown in the API playground. The result image is hosted on the Vidgo CDN.

![Aurora Ridge Station Cutaway](https://cdn.vidgo.ai/apis/models/bytedance/seedream/v4.5/text-to-image/v1/01/output.jpg)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "PGQCZ3X31DC6OEXQ",
    "status": "finished",
    "files": [
      {
        "file_url": "https://cdn.vidgo.ai/apis/models/bytedance/seedream/v4.5/text-to-image/v1/01/output.jpg",
        "file_type": "image"
      }
    ],
    "created_time": "2026-09-25T16:14:03",
    "error_message": null,
    "progress": 100
  }
}
```


# Seedream 4 Edit
Source: https://docs.vidgo.ai/api-manual/image-series/bytedance-seedream-v4-edit

api-manual/image-series/bytedance-seedream-v4-edit.json POST /api/generate/submit
Transform images with Seedream 4 using multiple visual references, with batch variations and output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Transform images with Seedream 4 using multiple visual references, with batch variations and output up to 4K.

## Models

* **bytedance/seedream/v4/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-10 images in `image_urls`.
* Choose from `1K`, `2K`, `4K` output resolutions.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `1:1`, `3:4`, `4:3`, `16:9`, `9:16`, `3:2`, `2:3`, `21:9`. Default: `1:1`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`, `4K`. Default: `2K`.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `15`. Default: `1`.
* The number of `image_urls` plus `n` totals 15 or fewer.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Input image URLs use HTTP(S) and point to publicly accessible image files.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-5,000 characters**.
* Include at least one non-whitespace character.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-10 items. Each item must be an HTTP(S) URL.


# Seedream 4 Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/bytedance-seedream-v4-text-to-image

api-manual/image-series/bytedance-seedream-v4-text-to-image.json POST /api/generate/submit
Generate image sets with Seedream 4, with output up to 4K for illustrations, design concepts, and visual variations.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate image sets with Seedream 4, with output up to 4K for illustrations, design concepts, and visual variations.

## Models

* **bytedance/seedream/v4/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Choose from `1K`, `2K`, `4K` output resolutions.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `1:1`, `3:4`, `4:3`, `16:9`, `9:16`, `3:2`, `2:3`, `21:9`. Default: `1:1`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`, `4K`. Default: `2K`.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `15`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-5,000 characters**.
* Include at least one non-whitespace character.


# Seedream 5.0 Lite Edit
Source: https://docs.vidgo.ai/api-manual/image-series/bytedance-seedream-v5-lite-edit

api-manual/image-series/bytedance-seedream-v5-lite-edit.json POST /api/generate/submit
Remix existing visuals with Seedream 5.0 Lite using text instructions and multiple reference images, with support for batch output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Remix existing visuals with Seedream 5.0 Lite using text instructions and multiple reference images, with support for batch output.

## Models

* **bytedance/seedream/v5/lite/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-10 images in `image_urls`.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Default: `1:1`.
* Presets: `2K`, `3K`, `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `21:9`.
* For custom dimensions, supply an object with integer `width` and `height`; both must be positive.
* Custom dimensions also accept a `WIDTHxHEIGHT` string, using `x` or `X` between the positive dimensions.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `15`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Input image URLs use HTTP(S) and point to publicly accessible JPEG or PNG files.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **3-3,000 characters**.
* Length is counted after leading and trailing whitespace is trimmed.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-10 items. Each item must be an HTTP(S) URL pointing to a JPEG or PNG image.

### Safety Checker

* `enable_safety_checker`: Optional. Set `true` to enable the content safety check or `false` to disable it. Default: `true`.


# Seedream 5.0 Lite Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/bytedance-seedream-v5-lite-text-to-image

api-manual/image-series/bytedance-seedream-v5-lite-text-to-image.json POST /api/generate/submit
Explore image ideas with Seedream 5.0 Lite, generating batches of visual variations from a written description.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Explore image ideas with Seedream 5.0 Lite, generating batches of visual variations from a written description.

## Models

* **bytedance/seedream/v5/lite/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Default: `1:1`.
* Presets: `2K`, `3K`, `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `3:2`, `2:3`, `21:9`.
* For custom dimensions, supply an object with integer `width` and `height`; both must be positive.
* Custom dimensions also accept a `WIDTHxHEIGHT` string, using `x` or `X` between the positive dimensions.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `15`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **3-3,000 characters**.
* Length is counted after leading and trailing whitespace is trimmed.

### Safety Checker

* `enable_safety_checker`: Optional. Set `true` to enable the content safety check or `false` to disable it. Default: `true`.


# Flux.2
Source: https://docs.vidgo.ai/api-manual/image-series/flux-2

api-manual/image-series/flux-2.json POST /api/generate/submit
32B parameter image generation model from Black Forest Labs with multi-reference support and superior text rendering

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

FLUX.2 is a 32 billion parameter image generation and editing model from Black Forest Labs. It combines text-to-image and multi-image editing in a single architecture, delivering photoreal images with clean typography at resolutions up to 2K. Supports referencing up to 8 images simultaneously with excellent character, product, and style consistency.

## Models

* **flux-2-pro** - High-fidelity text-to-image generation for production deployments
* **flux-2-pro-edit** - Advanced multi-reference image editing with up to 8 input images
* **flux-2-flex** - Adjustable speed vs. quality balance for flexible workflows
* **flux-2-flex-edit** - Flexible image editing with multi-reference support

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [blackforestlabs/flux-2-pro/text-to-image](/api-manual/image-series/blackforestlabs-flux-2-pro-text-to-image)
* [blackforestlabs/flux-2-pro/edit](/api-manual/image-series/blackforestlabs-flux-2-pro-edit)
* [blackforestlabs/flux-2-flex/text-to-image](/api-manual/image-series/blackforestlabs-flux-2-flex-text-to-image)
* [blackforestlabs/flux-2-flex/edit](/api-manual/image-series/blackforestlabs-flux-2-flex-edit)


# Nano Banana 2 Lite Edit
Source: https://docs.vidgo.ai/api-manual/image-series/google-nano-banana-2-lite-edit

api-manual/image-series/google-nano-banana-2-lite-edit.json POST /api/generate/submit
Combine text instructions with up to fourteen reference images in Nano Banana 2 Lite to explore new scenes and visual variations.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Combine text instructions with up to fourteen reference images in Nano Banana 2 Lite to explore new scenes and visual variations.

## Models

* **google/nano-banana-2-lite/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-14 images in `image_urls`.

## Output Options

### Aspect Ratio

* `size`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `3:4`, `4:3`, `4:5`, `5:4`, `9:16`, `16:9`, `21:9`. If omitted, no explicit aspect ratio is specified; there is no default. `auto` is not accepted.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes. Minimum length: **1 character**.
* The value must not be blank when supplied. No maximum prompt length is specified.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-14 items. Each item must be an HTTP(S) URL.

## Billing Factors

Billed per generation. Text-to-image and image editing use the same billing rate. The selected aspect ratio (`size`) and reference image count do not add charges.

Credits are deducted on submission and refunded if the task fails.


# Nano Banana 2 Lite Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/google-nano-banana-2-lite-text-to-image

api-manual/image-series/google-nano-banana-2-lite-text-to-image.json POST /api/generate/submit
Turn visual ideas into images with Nano Banana 2 Lite, with framing options for portraits, landscapes, and wide compositions.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn visual ideas into images with Nano Banana 2 Lite, with framing options for portraits, landscapes, and wide compositions.

## Models

* **google/nano-banana-2-lite/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.

## Output Options

### Aspect Ratio

* `size`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `3:4`, `4:3`, `4:5`, `5:4`, `9:16`, `16:9`, `21:9`. If omitted, no explicit aspect ratio is specified; there is no default. `auto` is not accepted.

For image editing, use `google/nano-banana-2-lite/edit`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the desired image. Minimum length: **1 character**.
* The value must not be blank when supplied. No maximum prompt length is specified.

## Billing Factors

Billed per generation. Text-to-image and image editing use the same billing rate. The selected aspect ratio (`size`) and reference image count do not add charges.

Credits are deducted on submission and refunded if the task fails.


# Nano Banana Edit
Source: https://docs.vidgo.ai/api-manual/image-series/google-nano-banana-edit

api-manual/image-series/google-nano-banana-edit.json POST /api/generate/submit
Use Nano Banana to reinterpret existing images through text instructions and visual references for new scenes, styles, and compositions.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Use Nano Banana to reinterpret existing images through text instructions and visual references for new scenes, styles, and compositions.

## Models

* **google/nano-banana/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts at least 1 image in `image_urls`.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `3:4`, `4:3`, `4:5`, `5:4`, `9:16`, `16:9`, `21:9`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-5,000 characters**.
* Include at least one non-whitespace character.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply at least 1 item. Each item must be an HTTP(S) URL.


# Nano Banana Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/google-nano-banana-text-to-image

api-manual/image-series/google-nano-banana-text-to-image.json POST /api/generate/submit
Create images from natural language with Nano Banana, exploring subjects and styles in square, portrait, or landscape compositions.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create images from natural language with Nano Banana, exploring subjects and styles in square, portrait, or landscape compositions.

## Models

* **google/nano-banana/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `3:4`, `4:3`, `4:5`, `5:4`, `9:16`, `16:9`, `21:9`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-5,000 characters**.
* Include at least one non-whitespace character.


# GPT-4o Image
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-4o-image

api-manual/image-series/gpt-4o-image.json POST /api/generate/submit
High-quality image generation and editing with GPT-4o

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate high-quality images using OpenAI's GPT-4o model. Supports text-to-image, image-to-image, and advanced image editing capabilities.

## Models

* **gpt-4o-image** - Text-to-image and image-to-image generation
* **gpt-4o-image-edit** - Advanced image editing with mask support

Generation with `gpt-4o-image` uses `n` to select and bill the requested image count. Editing with `gpt-4o-image-edit` produces and bills one image per request; `n` does not increase editing outputs.


# GPT Image 2
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2

api-manual/image-series/gpt-image-2.json POST /api/generate/submit
Generate and edit images with GPT Image 2

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **gpt-image-2** - Text-to-image generation from a prompt.
* **gpt-image-2-edit** - Image editing based on one or more `image_urls` and a text instruction.

## Overview

Use Vidgo API to generate a single image from a prompt or edit one or more reference images with natural-language instructions. GPT Image 2 supports explicit quality tiers, ratio presets, custom dimensions, and `1K`, `2K`, or `4K` resolution control.

## Resolution

The `resolution` parameter controls the effective output resolution.

| Resolution | Description |
| - | - |
| `1K` | Standard resolution |
| `2K` | High resolution |
| `4K` | Ultra-high resolution for true 4K-supported sizes |

### Billing Factors

Each request generates one image. Billing depends on `quality` and the effective output resolution determined by `size` and `resolution`. The rules below explain how the effective resolution is selected.

## Resolution Rules

* If `size` is omitted or set to `auto`, the request is processed and billed as `1K`, even if `resolution` is provided.
* Custom `WIDTHxHEIGHT` sizes require `resolution` `2K` or `4K`.
* `4K` billing applies only to `16:9`, `9:16`, `21:9`, or custom sizes with a `3840`-pixel edge.
* A `4K` request that does not meet the true 4K rule is billed at the effective `2K` tier.
* A custom size whose longest edge is `3840` must use `resolution: "4K"`.

## Custom Size Constraints

Custom sizes use `WIDTHxHEIGHT` format, for example `2304x1536`, and must satisfy all of the following:

* Width and height must both be divisible by `16`.
* The maximum edge length is `3840` pixels.
* The aspect ratio must not exceed `3:1`.
* Total pixel count must stay between `655,360` and `8,294,400`.

## Request Examples

### Generate Image

```bash theme={null}
curl --request POST \
  --url https://api.vidgo.ai/api/generate/submit \
  --header "Authorization: Bearer VIDGO_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
    "model": "gpt-image-2",
    "callback_url": "https://your-domain.com/callback",
    "input": {
      "prompt": "A premium product photo of a silver espresso machine on a clean white studio background, realistic lighting, high detail",
      "quality": "low",
      "size": "1:1",
      "resolution": "1K"
    }
  }'
```

### Edit Image

```bash theme={null}
curl --request POST \
  --url https://api.vidgo.ai/api/generate/submit \
  --header "Authorization: Bearer VIDGO_API_KEY" \
  --header "Content-Type: application/json" \
  --data '{
    "model": "gpt-image-2-edit",
    "callback_url": "https://your-domain.com/callback",
    "input": {
      "prompt": "Replace the background with a clean white studio backdrop and add a soft natural shadow while preserving the product shape",
      "quality": "medium",
      "size": "1:1",
      "resolution": "2K",
      "image_urls": [
        "https://cdn.vidgo.ai/temp/reference-product.jpg"
      ]
    }
  }'
```

## Response Flow

The submit endpoint returns a `task_id`, `status`, and `created_time`. If you provide `callback_url`, Vidgo API sends a POST request when the task reaches `finished` or `failed`.

Use the unified status endpoint to retrieve the final result:

```bash theme={null}
curl --request GET \
  --url https://api.vidgo.ai/api/generate/status/task-unified-1757165031-uyujaw3d \
  --header "Authorization: Bearer VIDGO_API_KEY"
```

When the task is `finished`, the status response includes generated image file URLs in `data.files`. If the task fails, the status payload returns the failure details for troubleshooting.

## Notes

* `input.prompt` is required and supports up to `4000` characters.
* `input.quality` is optional: `low`, `medium`, or `high`. Default: `low`.
* `input.size` is optional. Supported values are `auto`, `1:1`, `2:3`, `3:2`, `3:4`, `4:3`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`, or a custom `WIDTHxHEIGHT` value.
* `input.resolution` is optional: `1K`, `2K`, or `4K`.
* `gpt-image-2-edit` requires `input.image_urls`.
* `gpt-image-2` does not accept `input.image_urls`.
* Each request returns one image.


# GPT Image 2.5 Flare Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2-5-flare-edit

api-manual/image-series/gpt-image-2-5-flare-edit.json POST /api/generate/submit
The GPT Image 2.5 Flare Image Edit API edits 1–16 reference images into one 1K–4K image, with optional masks, custom sizes and transparent backgrounds.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

GPT Image 2.5 Flare Image Edit edits images from text instructions: send 1–16 reference images to restyle, replace backgrounds, combine products into a new scene, or change only a masked region. Flare is the faster GPT Image 2.5 option, suited to quick iteration.

For prompt-only generation, use [GPT Image 2.5 Flare Text to Image](/api-manual/image-series/gpt-image-2-5-flare-text-to-image) (`openai/gpt-image-2.5-flare/text-to-image`).

## Models

* **openai/gpt-image-2.5-flare/edit** — Image edit with GPT Image 2.5 Flare, 1–16 reference images, one image per request.

## Features

* Instruction-based editing with up to 16 reference images, referenced by order in the prompt.
* Optional mask to limit the edit to one region.
* Three resolution tiers (`1K`, `2K`, `4K`) with aspect-ratio presets, `auto`, or custom `WIDTHxHEIGHT` sizes.
* Five quality tiers: `low`, `medium`, `high`, `xhigh`, `max`.
* Transparent or opaque backgrounds; PNG, JPEG or WebP output.

## Required parameters

* `prompt` (string): Text instructions describing the edit. Up to 32,000 characters; must not be blank.
* `image_urls` (string\[]): 1–16 reference images as publicly accessible HTTP(S) URLs, in the order you refer to them in the prompt.

## Optional parameters

* `size` (string): Aspect-ratio preset, `auto`, or custom `WIDTHxHEIGHT`. When omitted, the output follows the first reference image's aspect ratio. See [Size](#size).
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`.
* `quality` (string): `low`, `medium`, `high`, `xhigh` or `max`. Default `high`.
* `background` (string): `auto`, `transparent` or `opaque`. Default `auto`. A transparent background requires `png` or `webp` output.
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.
* `mask_url` (string): Mask image selecting the area to edit, as a publicly accessible HTTP(S) URL. Requires `image_urls`.

## Size

* Presets: `1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`. When `size` is omitted, the output follows the first reference image's aspect ratio, using the closest supported ratio at the selected `resolution`. Send `size` to override it.
* Presets keep the selected `resolution` tier; actual pixel dimensions may vary between generations. Use a custom size when you need exact pixels.
* `auto` lets the model choose the size (about 1.5 megapixels). It works with any `resolution` and is billed at the `1K` tier.
* Custom `WIDTHxHEIGHT` (use `x` or `X`, for example `2304x1536`) works with any `resolution`. Both edges must be multiples of 16, the longest edge must not exceed 3,840 pixels, the aspect ratio must not exceed 3:1, and total pixels must be between 655,360 and 8,294,400.
* A custom size is billed at the tier that matches its total pixels, or at the selected `resolution` if that is higher:

| Total pixels (custom size) | Billing tier |
| - | - |
| Up to 1,572,864 | `1K` |
| 1,572,865–4,194,304 | `2K` |
| Above 4,194,304 | `4K` |

## Examples

* **Edit with a reference image**: replace the background at `2K` without `size`, so the output keeps the reference image's aspect ratio.
* **Mask edit**: change only the masked area, with `size: auto`.
* **Transparent background**: remove the background, WebP, `4K`, `xhigh` quality.
* **Custom size**: place the product in a panoramic scene at `2304x1536`, `2K`, `medium` quality.

## Notes

* Each request returns exactly one image; `n` other than `1` returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* A transparent background with `output_format: jpeg` returns 400.
* An empty `image_urls` array, more than 16 images, or a URL that is not a valid HTTP(S) URL returns 400.
* Credits depend on `quality`, the billing tier (`resolution`, or the custom-size tier described in [Size](#size)) and the number of reference images. The per-image maximum is reserved at submission; when the task finishes you are charged for actual token usage, up to that maximum, and the rest is refunded. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# GPT Image 2.5 Flare Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2-5-flare-text-to-image

api-manual/image-series/gpt-image-2-5-flare-text-to-image.json POST /api/generate/submit
The GPT Image 2.5 Flare Text to Image API turns a prompt into one 1K, 2K or 4K image, with preset or custom sizes, five quality tiers and transparent backgrounds.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

GPT Image 2.5 Flare Text to Image generates one image from a text prompt, from quick product shots to transparent assets and exact-pixel layouts at 1K, 2K or 4K. Flare is the faster GPT Image 2.5 option, suited to quick iteration.

For reference-image editing, use [GPT Image 2.5 Flare Image Edit](/api-manual/image-series/gpt-image-2-5-flare-edit) (`openai/gpt-image-2.5-flare/edit`).

## Models

* **openai/gpt-image-2.5-flare/text-to-image** — Text to image with GPT Image 2.5 Flare, one image per request.

## Features

* Text-to-image from a prompt of up to 32,000 characters.
* Three resolution tiers (`1K`, `2K`, `4K`) with aspect-ratio presets, `auto`, or custom `WIDTHxHEIGHT` sizes.
* Five quality tiers: `low`, `medium`, `high`, `xhigh`, `max`.
* Transparent or opaque backgrounds; PNG, JPEG or WebP output.

## Required parameters

* `prompt` (string): Text describing the desired image. Up to 32,000 characters; must not be blank.

## Optional parameters

* `size` (string): Aspect-ratio preset, `auto`, or custom `WIDTHxHEIGHT`. Default `1:1`. See [Size](#size).
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`.
* `quality` (string): `low`, `medium`, `high`, `xhigh` or `max`. Default `high`.
* `background` (string): `auto`, `transparent` or `opaque`. Default `auto`. A transparent background requires `png` or `webp` output.
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.

## Size

* Presets: `1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`. When `size` is omitted, the output is `1:1` at the selected `resolution`.
* Presets keep the selected `resolution` tier; actual pixel dimensions may vary between generations. Use a custom size when you need exact pixels.
* `auto` lets the model choose the size (about 1.5 megapixels). It works with any `resolution` and is billed at the `1K` tier.
* Custom `WIDTHxHEIGHT` (use `x` or `X`, for example `2304x1536`) works with any `resolution`. Both edges must be multiples of 16, the longest edge must not exceed 3,840 pixels, the aspect ratio must not exceed 3:1, and total pixels must be between 655,360 and 8,294,400.
* A custom size is billed at the tier that matches its total pixels, or at the selected `resolution` if that is higher:

| Total pixels (custom size) | Billing tier |
| - | - |
| Up to 1,572,864 | `1K` |
| 1,572,865–4,194,304 | `2K` |
| Above 4,194,304 | `4K` |

## Examples

* **Generate one image**: a studio product photo, `1:1`, `1K`, with a `callback_url`.
* **Transparent asset**: a cutout with `background: transparent`, WebP, `4K`, `xhigh` quality.
* **Custom size**: a panoramic scene at `2304x1536`, `2K`, `medium` quality.

## Notes

* Each request returns exactly one image; `n` other than `1` returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* A transparent background with `output_format: jpeg` returns 400.
* Credits depend on `quality` and the billing tier (`resolution`, or the custom-size tier described in [Size](#size)). The per-image maximum is reserved at submission; when the task finishes you are charged for actual token usage, up to that maximum, and the rest is refunded. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# GPT Image 2.5 Sunburst Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2-5-sunburst-edit

api-manual/image-series/gpt-image-2-5-sunburst-edit.json POST /api/generate/submit
The GPT Image 2.5 Sunburst Image Edit API edits 1–16 reference images into one 1K–4K image, with optional masks, custom sizes and transparent backgrounds.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

GPT Image 2.5 Sunburst Image Edit edits images from text instructions: send 1–16 reference images to restyle, replace backgrounds, combine products into a new scene, or change only a masked region. Sunburst is the GPT Image 2.5 option for detailed visual work and precision-focused editing.

For prompt-only generation, use [GPT Image 2.5 Sunburst Text to Image](/api-manual/image-series/gpt-image-2-5-sunburst-text-to-image) (`openai/gpt-image-2.5-sunburst/text-to-image`).

## Models

* **openai/gpt-image-2.5-sunburst/edit** — Image edit with GPT Image 2.5 Sunburst, 1–16 reference images, one image per request.

## Features

* Instruction-based editing with up to 16 reference images, referenced by order in the prompt.
* Optional mask to limit the edit to one region.
* Three resolution tiers (`1K`, `2K`, `4K`) with aspect-ratio presets, `auto`, or custom `WIDTHxHEIGHT` sizes.
* Five quality tiers: `low`, `medium`, `high`, `xhigh`, `max`.
* Transparent or opaque backgrounds; PNG, JPEG or WebP output.

## Required parameters

* `prompt` (string): Text instructions describing the edit. Up to 32,000 characters; must not be blank.
* `image_urls` (string\[]): 1–16 reference images as publicly accessible HTTP(S) URLs, in the order you refer to them in the prompt.

## Optional parameters

* `size` (string): Aspect-ratio preset, `auto`, or custom `WIDTHxHEIGHT`. When omitted, the output follows the first reference image's aspect ratio. See [Size](#size).
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`.
* `quality` (string): `low`, `medium`, `high`, `xhigh` or `max`. Default `high`.
* `background` (string): `auto`, `transparent` or `opaque`. Default `auto`. A transparent background requires `png` or `webp` output.
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.
* `mask_url` (string): Mask image selecting the area to edit, as a publicly accessible HTTP(S) URL. Requires `image_urls`.

## Size

* Presets: `1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`. When `size` is omitted, the output follows the first reference image's aspect ratio, using the closest supported ratio at the selected `resolution`. Send `size` to override it.
* Presets keep the selected `resolution` tier; actual pixel dimensions may vary between generations. Use a custom size when you need exact pixels.
* `auto` lets the model choose the size (about 1.5 megapixels). It works with any `resolution` and is billed at the `1K` tier.
* Custom `WIDTHxHEIGHT` (use `x` or `X`, for example `2304x1536`) works with any `resolution`. Both edges must be multiples of 16, the longest edge must not exceed 3,840 pixels, the aspect ratio must not exceed 3:1, and total pixels must be between 655,360 and 8,294,400.
* A custom size is billed at the tier that matches its total pixels, or at the selected `resolution` if that is higher:

| Total pixels (custom size) | Billing tier |
| - | - |
| Up to 1,572,864 | `1K` |
| 1,572,865–4,194,304 | `2K` |
| Above 4,194,304 | `4K` |

## Examples

* **Edit with a reference image**: replace the background at `2K` without `size`, so the output keeps the reference image's aspect ratio.
* **Mask edit**: change only the masked area, with `size: auto`.
* **Transparent background**: remove the background, WebP, `4K`, `xhigh` quality.
* **Custom size**: place the product in a panoramic scene at `2304x1536`, `2K`, `medium` quality.

## Notes

* Each request returns exactly one image; `n` other than `1` returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* A transparent background with `output_format: jpeg` returns 400.
* An empty `image_urls` array, more than 16 images, or a URL that is not a valid HTTP(S) URL returns 400.
* Credits depend on `quality`, the billing tier (`resolution`, or the custom-size tier described in [Size](#size)) and the number of reference images. The per-image maximum is reserved at submission; when the task finishes you are charged for actual token usage, up to that maximum, and the rest is refunded. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# GPT Image 2.5 Sunburst Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2-5-sunburst-text-to-image

api-manual/image-series/gpt-image-2-5-sunburst-text-to-image.json POST /api/generate/submit
The GPT Image 2.5 Sunburst Text to Image API turns a prompt into one 1K, 2K or 4K image, with preset or custom sizes, five quality tiers and transparent backgrounds.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

GPT Image 2.5 Sunburst Text to Image generates one image from a text prompt, from quick product shots to transparent assets and exact-pixel layouts at 1K, 2K or 4K. Sunburst is the GPT Image 2.5 option for detailed visual work and precision-focused editing.

For reference-image editing, use [GPT Image 2.5 Sunburst Image Edit](/api-manual/image-series/gpt-image-2-5-sunburst-edit) (`openai/gpt-image-2.5-sunburst/edit`).

## Models

* **openai/gpt-image-2.5-sunburst/text-to-image** — Text to image with GPT Image 2.5 Sunburst, one image per request.

## Features

* Text-to-image from a prompt of up to 32,000 characters.
* Three resolution tiers (`1K`, `2K`, `4K`) with aspect-ratio presets, `auto`, or custom `WIDTHxHEIGHT` sizes.
* Five quality tiers: `low`, `medium`, `high`, `xhigh`, `max`.
* Transparent or opaque backgrounds; PNG, JPEG or WebP output.

## Required parameters

* `prompt` (string): Text describing the desired image. Up to 32,000 characters; must not be blank.

## Optional parameters

* `size` (string): Aspect-ratio preset, `auto`, or custom `WIDTHxHEIGHT`. Default `1:1`. See [Size](#size).
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`.
* `quality` (string): `low`, `medium`, `high`, `xhigh` or `max`. Default `high`.
* `background` (string): `auto`, `transparent` or `opaque`. Default `auto`. A transparent background requires `png` or `webp` output.
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.

## Size

* Presets: `1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`. When `size` is omitted, the output is `1:1` at the selected `resolution`.
* Presets keep the selected `resolution` tier; actual pixel dimensions may vary between generations. Use a custom size when you need exact pixels.
* `auto` lets the model choose the size (about 1.5 megapixels). It works with any `resolution` and is billed at the `1K` tier.
* Custom `WIDTHxHEIGHT` (use `x` or `X`, for example `2304x1536`) works with any `resolution`. Both edges must be multiples of 16, the longest edge must not exceed 3,840 pixels, the aspect ratio must not exceed 3:1, and total pixels must be between 655,360 and 8,294,400.
* A custom size is billed at the tier that matches its total pixels, or at the selected `resolution` if that is higher:

| Total pixels (custom size) | Billing tier |
| - | - |
| Up to 1,572,864 | `1K` |
| 1,572,865–4,194,304 | `2K` |
| Above 4,194,304 | `4K` |

## Examples

* **Generate one image**: a studio product photo, `1:1`, `1K`, with a `callback_url`.
* **Transparent asset**: a cutout with `background: transparent`, WebP, `4K`, `xhigh` quality.
* **Custom size**: a panoramic scene at `2304x1536`, `2K`, `medium` quality.

## Notes

* Each request returns exactly one image; `n` other than `1` returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* A transparent background with `output_format: jpeg` returns 400.
* Credits depend on `quality` and the billing tier (`resolution`, or the custom-size tier described in [Size](#size)). The per-image maximum is reserved at submission; when the task finishes you are charged for actual token usage, up to that maximum, and the rest is refunded. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# GPT Image 2.6 Flare Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2-6-flare-edit

api-manual/image-series/gpt-image-2-6-flare-edit.json POST /api/generate/submit
The GPT Image 2.6 Flare Image Edit API edits 1–16 reference images into one 1K–4K image, with optional masks, transparent backgrounds and output compression.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

GPT Image 2.6 Flare Image Edit edits images from text instructions: send 1–16 reference images to restyle, replace backgrounds, combine products into a new scene, or change only a masked region. Flare is the faster GPT Image 2.6 option, suited to quick iteration.

For prompt-only generation, use [GPT Image 2.6 Flare Text to Image](/api-manual/image-series/gpt-image-2-6-flare-text-to-image) (`openai/gpt-image-2.6-flare/text-to-image`). GPT Image 2.6 accepts the same parameters as [GPT Image 2.5 Flare](/api-manual/image-series/gpt-image-2-5-flare-edit), plus `output_compression`, `moderation` and `complexity`.

## Models

* **openai/gpt-image-2.6-flare/edit** — Image edit with GPT Image 2.6 Flare, 1–16 reference images, one image per request.

## Features

* Instruction-based editing with up to 16 reference images, referenced by order in the prompt.
* Optional mask to limit the edit to one region.
* Three resolution tiers (`1K`, `2K`, `4K`) with aspect-ratio presets, `auto`, or custom `WIDTHxHEIGHT` sizes.
* Five quality tiers: `low`, `medium`, `high`, `xhigh`, `max`.
* Transparent or opaque backgrounds; PNG, JPEG or WebP output.
* Adjustable JPEG/WebP compression, moderation strictness and planning effort (`complexity`).

## Required parameters

* `prompt` (string): Text instructions describing the edit. Up to 32,000 characters; must not be blank.
* `image_urls` (string\[]): 1–16 reference images as publicly accessible HTTP(S) URLs, in the order you refer to them in the prompt.

## Optional parameters

* `size` (string): Aspect-ratio preset, `auto`, or custom `WIDTHxHEIGHT`. When omitted, the output follows the first reference image's aspect ratio. See [Size](#size).
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`.
* `quality` (string): `low`, `medium`, `high`, `xhigh` or `max`. Default `high`.
* `background` (string): `auto`, `transparent` or `opaque`. Default `auto`. A transparent background requires `png` or `webp` output.
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.
* `output_compression` (integer): Compression level from `0` to `100`. Only allowed when `output_format` is `jpeg` or `webp`. When omitted, the default compression applies.
* `mask_url` (string): Mask image selecting the area to edit, as a publicly accessible HTTP(S) URL. Requires `image_urls`.
* `moderation` (string): Content moderation strictness, `auto` or `low`. Default `auto`; `low` applies less restrictive filtering.
* `complexity` (string): Planning effort the model spends before generating, `low`, `medium` or `high`. Default `high`. Does not change the price.

## Size

* Presets: `1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`. When `size` is omitted, the output follows the first reference image's aspect ratio, using the closest supported ratio at the selected `resolution`. Send `size` to override it.
* Presets keep the selected `resolution` tier; actual pixel dimensions may vary between generations. Use a custom size when you need exact pixels.
* `auto` lets the model choose the size (about 1.5 megapixels). It works with any `resolution` and is billed at the `1K` tier.
* Custom `WIDTHxHEIGHT` (use `x` or `X`, for example `2304x1536`) works with any `resolution`. Both edges must be multiples of 16, the longest edge must not exceed 3,840 pixels, the aspect ratio must not exceed 3:1, and total pixels must be between 655,360 and 8,294,400.
* A custom size is billed at the tier that matches its total pixels, or at the selected `resolution` if that is higher:

| Total pixels (custom size) | Billing tier |
| - | - |
| Up to 1,572,864 | `1K` |
| 1,572,865–4,194,304 | `2K` |
| Above 4,194,304 | `4K` |

## Examples

* **Edit with a reference image**: replace the background at `2K` without `size`, so the output keeps the reference image's aspect ratio.
* **Mask edit**: change only the masked area, with `size: auto`.
* **Transparent background**: remove the background, WebP, `4K`, `xhigh` quality.
* **Custom size**: place the product in a panoramic scene at `2304x1536`, `2K`, `medium` quality.
* **Compressed JPEG**: `3:2`, `medium` quality, `output_format: jpeg`, `output_compression: 80`.

## Notes

* Each request returns exactly one image; `n` other than `1` returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* A transparent background with `output_format: jpeg` returns 400. `output_compression` with `output_format: png` also returns 400.
* An empty `image_urls` array, more than 16 images, or a URL that is not a valid HTTP(S) URL returns 400.
* Credits depend on `quality`, the billing tier (`resolution`, or the custom-size tier described in [Size](#size)) and the number of reference images. `complexity` and `moderation` do not change the price. The per-image maximum is reserved at submission; when the task finishes you are charged for actual token usage, up to that maximum, and the rest is refunded. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# GPT Image 2.6 Flare Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2-6-flare-text-to-image

api-manual/image-series/gpt-image-2-6-flare-text-to-image.json POST /api/generate/submit
The GPT Image 2.6 Flare Text to Image API turns a prompt into one 1K, 2K or 4K image, with transparent backgrounds, output compression and moderation control.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

GPT Image 2.6 Flare Text to Image generates one image from a text prompt, from quick product shots to transparent assets and exact-pixel layouts at 1K, 2K or 4K. Flare is the faster GPT Image 2.6 option, suited to quick iteration.

For reference-image editing, use [GPT Image 2.6 Flare Image Edit](/api-manual/image-series/gpt-image-2-6-flare-edit) (`openai/gpt-image-2.6-flare/edit`). GPT Image 2.6 accepts the same parameters as [GPT Image 2.5 Flare](/api-manual/image-series/gpt-image-2-5-flare-text-to-image), plus `output_compression`, `moderation` and `complexity`.

## Models

* **openai/gpt-image-2.6-flare/text-to-image** — Text to image with GPT Image 2.6 Flare, one image per request.

## Features

* Text-to-image from a prompt of up to 32,000 characters.
* Three resolution tiers (`1K`, `2K`, `4K`) with aspect-ratio presets, `auto`, or custom `WIDTHxHEIGHT` sizes.
* Five quality tiers: `low`, `medium`, `high`, `xhigh`, `max`.
* Transparent or opaque backgrounds; PNG, JPEG or WebP output.
* Adjustable JPEG/WebP compression, moderation strictness and planning effort (`complexity`).

## Required parameters

* `prompt` (string): Text describing the desired image. Up to 32,000 characters; must not be blank.

## Optional parameters

* `size` (string): Aspect-ratio preset, `auto`, or custom `WIDTHxHEIGHT`. Default `1:1`. See [Size](#size).
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`.
* `quality` (string): `low`, `medium`, `high`, `xhigh` or `max`. Default `high`.
* `background` (string): `auto`, `transparent` or `opaque`. Default `auto`. A transparent background requires `png` or `webp` output.
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.
* `output_compression` (integer): Compression level from `0` to `100`. Only allowed when `output_format` is `jpeg` or `webp`. When omitted, the default compression applies.
* `moderation` (string): Content moderation strictness, `auto` or `low`. Default `auto`; `low` applies less restrictive filtering.
* `complexity` (string): Planning effort the model spends before generating, `low`, `medium` or `high`. Default `high`. Does not change the price.

## Size

* Presets: `1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`. When `size` is omitted, the output is `1:1` at the selected `resolution`.
* Presets keep the selected `resolution` tier; actual pixel dimensions may vary between generations. Use a custom size when you need exact pixels.
* `auto` lets the model choose the size (about 1.5 megapixels). It works with any `resolution` and is billed at the `1K` tier.
* Custom `WIDTHxHEIGHT` (use `x` or `X`, for example `2304x1536`) works with any `resolution`. Both edges must be multiples of 16, the longest edge must not exceed 3,840 pixels, the aspect ratio must not exceed 3:1, and total pixels must be between 655,360 and 8,294,400.
* A custom size is billed at the tier that matches its total pixels, or at the selected `resolution` if that is higher:

| Total pixels (custom size) | Billing tier |
| - | - |
| Up to 1,572,864 | `1K` |
| 1,572,865–4,194,304 | `2K` |
| Above 4,194,304 | `4K` |

## Examples

* **Generate one image**: a studio product photo, `1:1`, `1K`, with a `callback_url`.
* **Transparent asset**: a cutout with `background: transparent`, WebP, `4K`, `xhigh` quality.
* **Custom size**: a panoramic scene at `2304x1536`, `2K`, `medium` quality.
* **Compressed JPEG**: `3:2`, `medium` quality, `output_format: jpeg`, `output_compression: 80`.

## Notes

* Each request returns exactly one image; `n` other than `1` returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* A transparent background with `output_format: jpeg` returns 400. `output_compression` with `output_format: png` also returns 400.
* Credits depend on `quality` and the billing tier (`resolution`, or the custom-size tier described in [Size](#size)). `complexity` and `moderation` do not change the price. The per-image maximum is reserved at submission; when the task finishes you are charged for actual token usage, up to that maximum, and the rest is refunded. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# GPT Image 2.6 Sunburst Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2-6-sunburst-edit

api-manual/image-series/gpt-image-2-6-sunburst-edit.json POST /api/generate/submit
The GPT Image 2.6 Sunburst Image Edit API edits 1–16 reference images into one 1K–4K image, with optional masks, transparent backgrounds and output compression.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

GPT Image 2.6 Sunburst Image Edit edits images from text instructions: send 1–16 reference images to restyle, replace backgrounds, combine products into a new scene, or change only a masked region. Sunburst is the GPT Image 2.6 option for detailed visual work and precision-focused editing.

For prompt-only generation, use [GPT Image 2.6 Sunburst Text to Image](/api-manual/image-series/gpt-image-2-6-sunburst-text-to-image) (`openai/gpt-image-2.6-sunburst/text-to-image`). GPT Image 2.6 accepts the same parameters as [GPT Image 2.5 Sunburst](/api-manual/image-series/gpt-image-2-5-sunburst-edit), plus `output_compression`, `moderation` and `complexity`.

## Models

* **openai/gpt-image-2.6-sunburst/edit** — Image edit with GPT Image 2.6 Sunburst, 1–16 reference images, one image per request.

## Features

* Instruction-based editing with up to 16 reference images, referenced by order in the prompt.
* Optional mask to limit the edit to one region.
* Three resolution tiers (`1K`, `2K`, `4K`) with aspect-ratio presets, `auto`, or custom `WIDTHxHEIGHT` sizes.
* Five quality tiers: `low`, `medium`, `high`, `xhigh`, `max`.
* Transparent or opaque backgrounds; PNG, JPEG or WebP output.
* Adjustable JPEG/WebP compression, moderation strictness and planning effort (`complexity`).

## Required parameters

* `prompt` (string): Text instructions describing the edit. Up to 32,000 characters; must not be blank.
* `image_urls` (string\[]): 1–16 reference images as publicly accessible HTTP(S) URLs, in the order you refer to them in the prompt.

## Optional parameters

* `size` (string): Aspect-ratio preset, `auto`, or custom `WIDTHxHEIGHT`. When omitted, the output follows the first reference image's aspect ratio. See [Size](#size).
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`.
* `quality` (string): `low`, `medium`, `high`, `xhigh` or `max`. Default `high`.
* `background` (string): `auto`, `transparent` or `opaque`. Default `auto`. A transparent background requires `png` or `webp` output.
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.
* `output_compression` (integer): Compression level from `0` to `100`. Only allowed when `output_format` is `jpeg` or `webp`. When omitted, the default compression applies.
* `mask_url` (string): Mask image selecting the area to edit, as a publicly accessible HTTP(S) URL. Requires `image_urls`.
* `moderation` (string): Content moderation strictness, `auto` or `low`. Default `auto`; `low` applies less restrictive filtering.
* `complexity` (string): Planning effort the model spends before generating, `low`, `medium` or `high`. Default `high`. Does not change the price.

## Size

* Presets: `1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`. When `size` is omitted, the output follows the first reference image's aspect ratio, using the closest supported ratio at the selected `resolution`. Send `size` to override it.
* Presets keep the selected `resolution` tier; actual pixel dimensions may vary between generations. Use a custom size when you need exact pixels.
* `auto` lets the model choose the size (about 1.5 megapixels). It works with any `resolution` and is billed at the `1K` tier.
* Custom `WIDTHxHEIGHT` (use `x` or `X`, for example `2304x1536`) works with any `resolution`. Both edges must be multiples of 16, the longest edge must not exceed 3,840 pixels, the aspect ratio must not exceed 3:1, and total pixels must be between 655,360 and 8,294,400.
* A custom size is billed at the tier that matches its total pixels, or at the selected `resolution` if that is higher:

| Total pixels (custom size) | Billing tier |
| - | - |
| Up to 1,572,864 | `1K` |
| 1,572,865–4,194,304 | `2K` |
| Above 4,194,304 | `4K` |

## Examples

* **Edit with a reference image**: replace the background at `2K` without `size`, so the output keeps the reference image's aspect ratio.
* **Mask edit**: change only the masked area, with `size: auto`.
* **Transparent background**: remove the background, WebP, `4K`, `xhigh` quality.
* **Custom size**: place the product in a panoramic scene at `2304x1536`, `2K`, `medium` quality.
* **Compressed JPEG**: `3:2`, `medium` quality, `output_format: jpeg`, `output_compression: 80`.

## Notes

* Each request returns exactly one image; `n` other than `1` returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* A transparent background with `output_format: jpeg` returns 400. `output_compression` with `output_format: png` also returns 400.
* An empty `image_urls` array, more than 16 images, or a URL that is not a valid HTTP(S) URL returns 400.
* Credits depend on `quality`, the billing tier (`resolution`, or the custom-size tier described in [Size](#size)) and the number of reference images. `complexity` and `moderation` do not change the price. The per-image maximum is reserved at submission; when the task finishes you are charged for actual token usage, up to that maximum, and the rest is refunded. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# GPT Image 2.6 Sunburst Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/gpt-image-2-6-sunburst-text-to-image

api-manual/image-series/gpt-image-2-6-sunburst-text-to-image.json POST /api/generate/submit
The GPT Image 2.6 Sunburst Text to Image API turns a prompt into one 1K, 2K or 4K image, with transparent backgrounds, output compression and moderation control.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

GPT Image 2.6 Sunburst Text to Image generates one image from a text prompt, from quick product shots to transparent assets and exact-pixel layouts at 1K, 2K or 4K. Sunburst is the GPT Image 2.6 option for detailed visual work and precision-focused editing.

For reference-image editing, use [GPT Image 2.6 Sunburst Image Edit](/api-manual/image-series/gpt-image-2-6-sunburst-edit) (`openai/gpt-image-2.6-sunburst/edit`). GPT Image 2.6 accepts the same parameters as [GPT Image 2.5 Sunburst](/api-manual/image-series/gpt-image-2-5-sunburst-text-to-image), plus `output_compression`, `moderation` and `complexity`.

## Models

* **openai/gpt-image-2.6-sunburst/text-to-image** — Text to image with GPT Image 2.6 Sunburst, one image per request.

## Features

* Text-to-image from a prompt of up to 32,000 characters.
* Three resolution tiers (`1K`, `2K`, `4K`) with aspect-ratio presets, `auto`, or custom `WIDTHxHEIGHT` sizes.
* Five quality tiers: `low`, `medium`, `high`, `xhigh`, `max`.
* Transparent or opaque backgrounds; PNG, JPEG or WebP output.
* Adjustable JPEG/WebP compression, moderation strictness and planning effort (`complexity`).

## Required parameters

* `prompt` (string): Text describing the desired image. Up to 32,000 characters; must not be blank.

## Optional parameters

* `size` (string): Aspect-ratio preset, `auto`, or custom `WIDTHxHEIGHT`. Default `1:1`. See [Size](#size).
* `resolution` (string): `1K`, `2K` or `4K`. Default `1K`.
* `quality` (string): `low`, `medium`, `high`, `xhigh` or `max`. Default `high`.
* `background` (string): `auto`, `transparent` or `opaque`. Default `auto`. A transparent background requires `png` or `webp` output.
* `output_format` (string): `png`, `jpeg` or `webp`. Default `png`.
* `output_compression` (integer): Compression level from `0` to `100`. Only allowed when `output_format` is `jpeg` or `webp`. When omitted, the default compression applies.
* `moderation` (string): Content moderation strictness, `auto` or `low`. Default `auto`; `low` applies less restrictive filtering.
* `complexity` (string): Planning effort the model spends before generating, `low`, `medium` or `high`. Default `high`. Does not change the price.

## Size

* Presets: `1:1`, `2:3`, `3:2`, `4:3`, `3:4`, `4:5`, `5:4`, `16:9`, `9:16`, `21:9`. When `size` is omitted, the output is `1:1` at the selected `resolution`.
* Presets keep the selected `resolution` tier; actual pixel dimensions may vary between generations. Use a custom size when you need exact pixels.
* `auto` lets the model choose the size (about 1.5 megapixels). It works with any `resolution` and is billed at the `1K` tier.
* Custom `WIDTHxHEIGHT` (use `x` or `X`, for example `2304x1536`) works with any `resolution`. Both edges must be multiples of 16, the longest edge must not exceed 3,840 pixels, the aspect ratio must not exceed 3:1, and total pixels must be between 655,360 and 8,294,400.
* A custom size is billed at the tier that matches its total pixels, or at the selected `resolution` if that is higher:

| Total pixels (custom size) | Billing tier |
| - | - |
| Up to 1,572,864 | `1K` |
| 1,572,865–4,194,304 | `2K` |
| Above 4,194,304 | `4K` |

## Examples

* **Generate one image**: a studio product photo, `1:1`, `1K`, with a `callback_url`.
* **Transparent asset**: a cutout with `background: transparent`, WebP, `4K`, `xhigh` quality.
* **Custom size**: a panoramic scene at `2304x1536`, `2K`, `medium` quality.
* **Compressed JPEG**: `3:2`, `medium` quality, `output_format: jpeg`, `output_compression: 80`.

## Notes

* Each request returns exactly one image; `n` other than `1` returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* A transparent background with `output_format: jpeg` returns 400. `output_compression` with `output_format: png` also returns 400.
* Credits depend on `quality` and the billing tier (`resolution`, or the custom-size tier described in [Size](#size)). `complexity` and `moderation` do not change the price. The per-image maximum is reserved at submission; when the task finishes you are charged for actual token usage, up to that maximum, and the rest is refunded. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Ideogram 4.5 Edit
Source: https://docs.vidgo.ai/api-manual/image-series/ideogram-v4-5-edit

api-manual/image-series/ideogram-v4-5-edit.json POST /api/generate/submit
The Ideogram 4.5 Edit API edits one image from a text instruction, with an optional mask, up to 4 reference images and regular or high edit precision.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Ideogram 4.5 Edit changes one existing image from a written instruction. Restrict the edit to a masked area, guide it with up to four reference images, and choose `high` edit precision when the untouched parts must stay pixel-exact. Typical uses are swapping products or signage text, retouching backgrounds and producing variants of a design. Ideogram says the model avoids the colour drift and artifacts that build up over repeated edits.

Ideogram 4.5 is Ideogram's image model for text-heavy visuals and precise edits. Vidgo exposes it as two model IDs: [Text to Image](/api-manual/image-series/ideogram-v4-5-text-to-image) and [Edit](/api-manual/image-series/ideogram-v4-5-edit).

## Models

* **ideogram/v4.5/edit** — Image edit with Ideogram 4.5: one source image, optional mask and references.

## Features

* Image edit from a natural-language instruction, including replacing text in the image.
* Optional mask to limit the edit to one area.
* Up to 4 reference images for style or content (3 when a mask is used).
* `edit_precision: high` for edits that keep everything else unchanged.
* Four quality tiers, including `very_low` for cheap drafts, and 1–8 results per request.

## Input modes

* **Instruction edit**: `image_urls` + `prompt`. The model decides which area to change.
* **Masked edit**: add `mask_url` (black = edit, white = keep) and keep `size: auto`.
* Either mode can add `reference_image_urls`.

## Required parameters

* `prompt` (string): What to change and how, 1–10,000 characters. Must not be blank.
* `image_urls` (string\[]): Exactly one image to edit. Public HTTP(S) URL or `data:image/...;base64,...` data URI.

## Optional parameters

* `mask_url` (string): Mask the same size as the source image, containing both black (edit) and white (keep). Requires `size: auto`.
* `reference_image_urls` (string\[]): Up to 4 reference images, or up to 3 with `mask_url`. They guide the result and are not edited.
* `edit_precision` (string): `regular` or `high`. Default `regular`. `high` requires `size: auto`.
* `quality` (string): `very_low`, `low`, `medium` or `high`. Default `medium`.
* `n` (integer): Number of results, `1`–`8`. Default `1`.
* `size` (string): Output size. Default `auto`. See [Size](#size).

## Size

* `auto` (default) keeps the source image's dimensions. It is required when you send `mask_url` or `edit_precision: high`.
* Presets: `1:1`, `1:1 HD`, `4:3`, `3:4`, `16:9`, `9:16`.
* Custom `WIDTHxHEIGHT`, for example `1536x1024`: each side at least 256 and a multiple of 32, at most 4,194,304 pixels in total (2048x2048), and an aspect ratio no wider than 6:1.

## Output

The finished task returns one file per requested image (`n`) in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Instruction-only edit**: change the text on a shop sign, `size: auto`, `medium` quality.
* **Masked edit with a reference and high precision**: a sofa mask plus an armchair reference image, `edit_precision: high`, `quality: high`, with a `callback_url`.

## Notes

* `image_urls` must contain exactly one image. To create an image from scratch, use [Ideogram 4.5 Text to Image](/api-manual/image-series/ideogram-v4-5-text-to-image).
* A `mask_url` or `edit_precision: high` with any `size` other than `auto` returns 400, and so do four references together with a mask.
* Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `quality` and `n` only; `size`, the mask and references do not change the price. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Ideogram 4.5 Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/ideogram-v4-5-text-to-image

api-manual/image-series/ideogram-v4-5-text-to-image.json POST /api/generate/submit
The Ideogram 4.5 Text to Image API generates 1–8 posters, logos and text-rich images per request, with low, medium or high quality and 42 sizes.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Ideogram 4.5 Text to Image turns a prompt into posters, logos, packaging mockups, social graphics and other images where legible lettering matters. Request up to eight images at once and pick a quality tier that fits a draft or a final asset.

Ideogram 4.5 is Ideogram's image model for text-heavy visuals and precise edits. Vidgo exposes it as two model IDs: [Text to Image](/api-manual/image-series/ideogram-v4-5-text-to-image) and [Edit](/api-manual/image-series/ideogram-v4-5-edit).

## Models

* **ideogram/v4.5/text-to-image** — Text-to-image with Ideogram 4.5, 1–8 images per request.

## Features

* Text-to-image focused on accurate in-image text (titles, slogans, labels).
* 1–8 images per request with `n`.
* Three quality tiers: `low`, `medium`, `high`.
* Six aspect-ratio presets plus 36 exact pixel sizes, from 720x1280 up to 3328x1248 and 2048x2048.

## Required parameters

* `prompt` (string): The image you want, 1–10,000 characters. Quote any text that must appear, for example `"GRAND OPENING"`. Must not be blank.

## Optional parameters

* `n` (integer): Number of images, `1`–`8`. Default `1`.
* `quality` (string): `low`, `medium` or `high`. Default `medium`.
* `size` (string): Output size. Default `1:1 HD`. See [Size](#size).

## Size

* Presets: `1:1`, `1:1 HD`, `4:3`, `3:4`, `16:9`, `9:16`. `1:1 HD` is a higher-resolution square.
* Exact pixel sizes: `720x1280`, `720x1440`, `800x1280`, `832x1248`, `864x1152`, `896x1120`, `1024x1024`, `1024x3072`, `1120x896`, `1152x864`, `1152x2944`, `1248x832`, `1248x3328`, `1280x720`, `1280x800`, `1280x3072`, `1296x3168`, `1440x720`, `1440x2560`, `1440x2880`, `1600x2560`, `1664x2496`, `1728x2304`, `1792x2240`, `2048x2048`, `2240x1792`, `2304x1728`, `2496x1664`, `2560x1440`, `2560x1600`, `2880x1440`, `2944x1152`, `3072x1024`, `3072x1280`, `3168x1296`, `3328x1248`.
* Any other value returns 400.

## Output

The finished task returns one file per requested image (`n`) in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Poster with lettering**: a travel poster whose title is quoted in the prompt, `3:4`, `medium` quality.
* **Four high-quality logo options with callback**: `n: 4`, `quality: high`, `1024x1024`, with a `callback_url`.

## Notes

* This ID takes text only. To change an existing image, use [Ideogram 4.5 Edit](/api-manual/image-series/ideogram-v4-5-edit).
* Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `quality` and `n` only; `size` does not change the price. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Ideogram Vector 1.0 Image to SVG
Source: https://docs.vidgo.ai/api-manual/image-series/ideogram-vector-1-0-image-to-vector

api-manual/image-series/ideogram-vector-1-0-image-to-vector.json POST /api/generate/submit
The Ideogram Vector 1.0 Image to SVG API converts one PNG, JPG or WebP image into an editable SVG file, in default (detailed) or simple mode.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Ideogram Vector 1.0 Image to SVG converts a raster image, such as a PNG logo, a scanned sketch or a JPG illustration, into an SVG file you can scale and edit in a vector tool. Choose `default` mode for detail or `simple` mode for a cleaner, flatter result. No prompt is needed.

Ideogram Vector 1.0 is Ideogram's first model that outputs vector graphics instead of pixels. Vidgo exposes it as two model IDs: [Text to SVG](/api-manual/image-series/ideogram-vector-1-0-text-to-vector) and [Image to SVG](/api-manual/image-series/ideogram-vector-1-0-image-to-vector).

## Models

* **ideogram/vector-1.0/image-to-vector** — Image to SVG with Ideogram Vector 1.0: one image in, one SVG out.

## Features

* Image-to-SVG vectorization of one image per request.
* Two styles: `default` (finer detail) and `simple` (fewer shapes).
* Accepts a public URL or an image data URI.

## Required parameters

* `image_urls` (string\[]): Exactly one image (PNG, JPG or WebP recommended). Public HTTP(S) URL or `data:image/...;base64,...` data URI.

## Optional parameters

* `mode` (string): `default` or `simple`. Default `default`.

## Output

The finished task returns SVG files (`image/svg+xml`) in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload. Open them in any vector editor or embed them directly in a web page.

## Examples

* **Vectorize a logo in default mode**: one PNG logo, `mode: default`.
* **Simple mode with callback**: a JPG sketch, `mode: simple`, with a `callback_url`.

## Notes

* This ID takes no `prompt`; sending one returns 400. To create new SVG artwork from text, use [Ideogram Vector 1.0 Text to SVG](/api-manual/image-series/ideogram-vector-1-0-text-to-vector).
* `image_urls` must contain exactly one image. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Each request is billed at one flat rate per SVG; `mode` does not change the price. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Ideogram Vector 1.0 Text to SVG
Source: https://docs.vidgo.ai/api-manual/image-series/ideogram-vector-1-0-text-to-vector

api-manual/image-series/ideogram-vector-1-0-text-to-vector.json POST /api/generate/submit
The Ideogram Vector 1.0 Text to SVG API creates 1–8 SVG files from a prompt, with up to 5 reference images, 18 aspect ratio options and three quality tiers.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Ideogram Vector 1.0 Text to SVG generates scalable SVG artwork from a prompt: logos, icons, stickers, badges and flat illustrations that stay sharp at any size. Lettering in the prompt is drawn as vector shapes. Add up to five reference images to steer style or composition.

Ideogram Vector 1.0 is Ideogram's first model that outputs vector graphics instead of pixels. Vidgo exposes it as two model IDs: [Text to SVG](/api-manual/image-series/ideogram-vector-1-0-text-to-vector) and [Image to SVG](/api-manual/image-series/ideogram-vector-1-0-image-to-vector).

## Models

* **ideogram/vector-1.0/text-to-vector** — Text to SVG with Ideogram Vector 1.0, 1–8 files per request.

## Features

* Text-to-SVG: native vector output (`image/svg+xml`), not traced raster images.
* Optional 1–5 reference images; the first one is the primary reference.
* 18 aspect ratio options, from tall `1:4` to wide `4:1`, plus `auto`.
* Three quality tiers and 1–8 SVG files per request.

## Required parameters

* `prompt` (string): The artwork you want, 1–9,500 characters. Quote any lettering, for example `"TRAILHEAD"`. Must not be blank.

## Optional parameters

* `image_urls` (string\[]): 1–5 reference images (public HTTP(S) URLs or image data URIs). Omit the field when you have none.
* `quality` (string): `low`, `medium` or `high`. Default `medium`.
* `n` (integer): Number of SVG files, `1`–`8`. Default `1`.
* `size` (string): Aspect ratio. Default `auto`. See [Size](#size).

## Size

`size` sets the aspect ratio, not a pixel size: `auto`, `1:1`, `1:4`, `1:3`, `1:2`, `9:16`, `10:16`, `2:3`, `3:4`, `4:5`, `5:4`, `4:3`, `3:2`, `16:10`, `16:9`, `2:1`, `3:1`, `4:1`.

## Output

The finished task returns SVG files (`image/svg+xml`) in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload. Open them in any vector editor or embed them directly in a web page.

## Examples

* **Prompt-only SVG icons**: two square icon sets at `medium` quality.
* **Banner with a style reference and callback**: a `3:1` banner with quoted lettering, one brand reference image, `quality: high`, with a `callback_url`.

## Notes

* To convert an existing raster image (PNG, JPG) into SVG, use [Ideogram Vector 1.0 Image to SVG](/api-manual/image-series/ideogram-vector-1-0-image-to-vector).
* An empty `image_urls` array or more than five references returns 400. Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `quality` and `n`. With `quality: high`, sending `image_urls` uses a higher per-file rate; `low` and `medium` cost the same with or without references. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Kling O1 Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/kwaivgi-kling-image-o1

api-manual/image-series/kwaivgi-kling-image-o1.json POST /api/generate/submit
Edit images with Kling O1 using visual references and subject elements, with batch output and resolutions up to 2K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Edit images with Kling O1 using visual references and subject elements, with batch output and resolutions up to 2K.

## Models

* **kwaivgi/kling-image-o1** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-10 images in `image_urls`.
* Choose from `1K`, `2K` output resolutions.
* Supply additional subject references through `elements`.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `auto`, `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `3:2`, `2:3`, `21:9`. Default: `auto`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`. Default: `1K`.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `9`. Default: `1`.

### Output Format

* `output_format`: Optional. Accepted values: `jpeg`, `png`, `webp`. Default: `png`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-2,000 characters**.
* The value must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-10 items. Each item must be an HTTP(S) URL. A single URL string and Base64 inputs are not accepted.

### Subject Elements

* `elements`: Optional array of subject-reference objects. Refer to them in the prompt as `@Element1`, `@Element2`, etc., in array order. Additional object properties are preserved. No element or additional-view count limit is specified.

* `elements[].frontal_image_url`: Optional. Use a publicly accessible HTTP(S) URL.

* `elements[].reference_image_urls`: Optional. Supply an array. Each item must be an HTTP(S) URL.

### Tested multi-image example

```json theme={null}
{
  "model": "kwaivgi/kling-image-o1",
  "input": {
    "prompt": "Use image 1 as the flooded forest boardwalk scene to edit and image 2 only as the animal identity reference. Edit the scene in image 1 by adding exactly one adult Malayan tapir, the tapir from image 2, standing on the wide wooden deck in the left-center foreground. Preserve the exact tapir identity and anatomy from image 2: black head, neck and hindquarters, broad white saddle, short flexible snout and white-rimmed ears. The tapir faces left in three-quarter side view, pausing as if listening toward the flooded forest. Its full body must be visible, four feet planted naturally on the boardwalk, correct real-world scale, weight and soft contact shadows. Preserve the boardwalk structure, tree positions, floodwater reflections, original camera perspective and documentary photographic style. Match the soft green canopy daylight. Do not add people, other animals, accessories, text, logos or watermark.",
    "image_urls": [
      "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-image-o1/edit/v1/01/input-01.png",
      "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-image-o1/edit/v1/01/input-02.png"
    ],
    "resolution": "2K",
    "size": "16:9",
    "output_format": "png",
    "n": 1
  }
}
```

## Billing Factors

Billing depends on output image count (`n`). Both supported resolutions use the same rate. Reference images and subject elements do not add charges.

Credits are deducted on submission and refunded if the task fails.


# Kling O3 Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/kwaivgi-kling-image-o3-edit

api-manual/image-series/kwaivgi-kling-image-o3-edit.json POST /api/generate/submit
Rework images with Kling O3 using multiple references and subject elements, with output up to 4K and support for visual variations.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Rework images with Kling O3 using multiple references and subject elements, with output up to 4K and support for visual variations.

## Models

* **kwaivgi/kling-image-o3/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-10 images in `image_urls`.
* Choose from `1K`, `2K`, `4K` output resolutions.
* Supply additional subject references through `elements`.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `auto`, `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `3:2`, `2:3`, `21:9`. Default: `16:9`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`, `4K`. Default: `1K`.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `9`. Default: `1`. Fractional values, strings, and booleans are rejected.

### Output Format

* `output_format`: Optional. Accepted values: `jpeg`, `png`, `webp`. Default: `png`.

## Advanced Parameters

Defaults apply only when an optional input field is omitted. Explicit `null` or empty-string values are rejected. An empty `elements` array is accepted.

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-2,000 characters**.
* The value must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-10 items. Each item must be an HTTP(S) URL.

### Subject Elements

* `elements`: Optional. Provide subject-reference objects. Supply an array. Each item is an object.

* `elements[].frontal_image_url`: Optional. Use a publicly accessible HTTP(S) URL.

* `elements[].reference_image_urls`: Optional. Supply an array. Each item must be an HTTP(S) URL.

Refer to subjects in the prompt as `@Element1`, `@Element2`, etc. Element objects may include additional fields; no element count limit is specified.

## Billing Factors

Billing depends on output resolution (`resolution`) and output image count (`n`). Text-to-image and image editing use the same rates. Reference images and subject elements do not add charges.

Credits are deducted on submission and refunded if the task fails.


# Kling O3 Image Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/kwaivgi-kling-image-o3-text-to-image

api-manual/image-series/kwaivgi-kling-image-o3-text-to-image.json POST /api/generate/submit
Generate images up to 4K with Kling O3, creating multiple variations in a range of compositions and output formats.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate images up to 4K with Kling O3, creating multiple variations in a range of compositions and output formats.

## Models

* **kwaivgi/kling-image-o3/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Choose from `1K`, `2K`, `4K` output resolutions.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`, `3:2`, `2:3`, `21:9`. Default: `16:9`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`, `4K`. Default: `1K`.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `9`. Default: `1`. Fractional values, strings, and booleans are rejected.

### Output Format

* `output_format`: Optional. Accepted values: `jpeg`, `png`, `webp`. Default: `png`.

## Advanced Parameters

Defaults apply only when an optional input field is omitted. Explicit `null` or empty-string values are rejected. An empty `elements` array is accepted.

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-2,000 characters**.
* The value must not be blank when supplied.

### Subject Elements

* `elements`: Optional array of subject-reference objects. Refer to them in the prompt as `@Element1`, `@Element2`, etc.
* `elements[].frontal_image_url`: Optional public HTTP(S) URL for the main subject view.
* `elements[].reference_image_urls`: Optional array of additional public HTTP(S) image URLs.
* Element objects may include additional fields. No element count limit is specified.

## Billing Factors

Billing depends on output resolution (`resolution`) and output image count (`n`). Text-to-image and image editing use the same rates. Reference images and subject elements do not add charges.

Credits are deducted on submission and refunded if the task fails.


# Nano Banana Pro
Source: https://docs.vidgo.ai/api-manual/image-series/nano-banana-2

api-manual/image-series/nano-banana-2.json POST /api/generate/submit
Next-generation AI image generation powered by Gemini 3 Pro Image Preview

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Next-generation AI image generation powered by Google's Gemini 3 Pro Image Preview. Delivers 2K native resolution images in under 1 second with enhanced text rendering, character consistency, and designer-grade quality.

## Models

* **nano-banana-2** - Text-to-image and image-to-image generation
* **nano-banana-2-edit** - Advanced image editing capabilities


# Nano Banana 2.1 Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/nano-banana-2-1-edit

api-manual/image-series/nano-banana-2-1-edit.json POST /api/generate/submit
Edit one image with Nano Banana 2.1 using text instructions and reference images.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `size` (string): Output aspect ratio. auto follows the reference image. Accepted values: `auto`, `1:1`, `2:3`, `3:2`, `3:4`, `4:3`, `4:5`, `5:4`, `9:16`, `16:9`, `21:9`. Default: `auto`.
* `resolution` (string): Output resolution tier. Accepted values: `1K`, `2K`, `4K`. Default: `1K`.

## Usage Guide

Edit one image with Nano Banana 2.1 using text instructions and reference images.

* Set `model` to `google/nano-banana-2.1/edit`.
* Place `model` and optional `callback_url` at the request root; place workflow parameters inside `input`.
* Each request produces one image.
* For prompt-only generation, use [Text to Image](/api-manual/image-series/nano-banana-2-1-text-to-image) with `model=google/nano-banana-2.1/text-to-image`.

## Parameter Details

* `prompt` is required. Supply non-blank text of up to 20,000 characters.
* `image_urls` is required and accepts 1 to 10 reference image URLs. Use publicly accessible HTTP(S) URLs in the same order as the references in your prompt.
* `size` defaults to `auto`. `auto` follows the reference image aspect ratio.

## Billing

Billing is per output image and depends on `resolution`. Text-to-image and editing use the same rate at each tier.

Credits are deducted on submission and refunded if the task fails.


# Nano Banana 2.1 Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/nano-banana-2-1-text-to-image

api-manual/image-series/nano-banana-2-1-text-to-image.json POST /api/generate/submit
Generate one image with Nano Banana 2.1 using a text prompt.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `size` (string): Output aspect ratio. Accepted values: `1:1`, `2:3`, `3:2`, `3:4`, `4:3`, `4:5`, `5:4`, `9:16`, `16:9`, `21:9`. Default: `1:1`.
* `resolution` (string): Output resolution tier. Accepted values: `1K`, `2K`, `4K`. Default: `1K`.

## Usage Guide

Generate one image with Nano Banana 2.1 using a text prompt.

* Set `model` to `google/nano-banana-2.1/text-to-image`.
* Place `model` and optional `callback_url` at the request root; place workflow parameters inside `input`.
* Each request produces one image.
* For reference-image editing, use [Image Edit](/api-manual/image-series/nano-banana-2-1-edit) with `model=google/nano-banana-2.1/edit`.

## Parameter Details

* `prompt` is required. Supply non-blank text of up to 20,000 characters.
* `size` defaults to `1:1`.

## Billing

Billing is per output image and depends on `resolution`. Text-to-image and editing use the same rate at each tier.

Credits are deducted on submission and refunded if the task fails.


# Nano Banana 2
Source: https://docs.vidgo.ai/api-manual/image-series/nano-banana-2-new

api-manual/image-series/nano-banana-2-new.json POST /api/generate/submit
Next-generation AI image generation powered by Gemini 3.1 Flash Image Preview

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Next-generation AI image generation powered by Google's Gemini 3.1 Flash Image Preview. Delivers native 2K / 4K resolution images with precise multilingual text rendering, chain-of-thought understanding of physical relationships, and support for up to 14 reference images.

## Models

* **nano-banana-2-new** - Text-to-image and image-to-image generation
* **nano-banana-2-new-edit** - Advanced image editing capabilities


# GPT Image 1.5 Edit
Source: https://docs.vidgo.ai/api-manual/image-series/openai-gpt-image-1-5-edit

api-manual/image-series/openai-gpt-image-1-5-edit.json POST /api/generate/submit
Make targeted image changes with GPT Image 1.5 using text instructions, reference images, and optional masks for selected areas.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Make targeted image changes with GPT Image 1.5 using text instructions, reference images, and optional masks for selected areas.

## Models

* **openai/gpt-image-1.5/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts at least 1 image in `image_urls`.
* Each editing request produces one image and is billed as one image.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `1024x1024`, `1024x1536`, `1536x1024`. Default: `1024x1024`.

### Image Count

* `n`: Optional compatibility parameter. Accepted values: `1`, `2`, `3`, `4`. Default: `1`. Each edit request produces one image and is billed for one image; larger values do not increase the output count.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-1,000 characters**.
* Include at least one non-whitespace character.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply at least 1 item. Each item must be an HTTP(S) URL.

### Mask URL

* `mask_url`: Optional. Publicly accessible HTTP(S) URL of a mask image that selects the area to edit.


# GPT Image 1.5 Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/openai-gpt-image-1-5-text-to-image

api-manual/image-series/openai-gpt-image-1-5-text-to-image.json POST /api/generate/submit
Turn written image concepts into square, portrait, or landscape visuals with GPT Image 1.5, with up to four variations per request.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn written image concepts into square, portrait, or landscape visuals with GPT Image 1.5, with up to four variations per request.

## Models

* **openai/gpt-image-1.5/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Control the number of output images with `n`.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `1024x1024`, `1024x1536`, `1536x1024`. Default: `1024x1024`.

### Image Count

* `n`: Optional. Number of images to generate. Accepted values: `1`, `2`, `3`, `4`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-1,000 characters**.
* Include at least one non-whitespace character.


# Image Models
Source: https://docs.vidgo.ai/api-manual/image-series/overview

All image generation and editing models available on Vidgo API, with their model IDs.

Every image model is called through `POST /api/generate/submit`. Send a model ID in `model`, the model's parameters in `input`, and an optional `callback_url`. The request returns a `task_id` right away and the task runs asynchronously.

## Workflow

1. Submit a task to `POST /api/generate/submit` with a model ID from the table below.
2. Poll [Get Task Status](/api-manual/task-management/status) with the returned `task_id`, or receive the result at `callback_url` (see [Webhook Callbacks](/api-manual/task-management/webhooks)).
3. Download the files in `files` when the status is `finished`.

Most models expose one model ID per mode, such as `text-to-image` and `edit`. Edit IDs take input images; check each model page for its inputs.

## Models

| Model | Provider | Model IDs |
| - | - | - |
| [GPT-4o Image](/api-manual/image-series/gpt-4o-image) | OpenAI | `gpt-4o-image` `gpt-4o-image-edit` |
| [GPT Image 1.5 Text to Image](/api-manual/image-series/openai-gpt-image-1-5-text-to-image) | OpenAI | `openai/gpt-image-1.5/text-to-image` |
| [GPT Image 1.5 Edit](/api-manual/image-series/openai-gpt-image-1-5-edit) | OpenAI | `openai/gpt-image-1.5/edit` |
| [GPT Image 2](/api-manual/image-series/gpt-image-2) | OpenAI | `gpt-image-2` `gpt-image-2-edit` |
| [GPT Image 2.5 Flare Text to Image](/api-manual/image-series/gpt-image-2-5-flare-text-to-image) | OpenAI | `openai/gpt-image-2.5-flare/text-to-image` |
| [GPT Image 2.5 Flare Image Edit](/api-manual/image-series/gpt-image-2-5-flare-edit) | OpenAI | `openai/gpt-image-2.5-flare/edit` |
| [GPT Image 2.5 Sunburst Text to Image](/api-manual/image-series/gpt-image-2-5-sunburst-text-to-image) | OpenAI | `openai/gpt-image-2.5-sunburst/text-to-image` |
| [GPT Image 2.5 Sunburst Image Edit](/api-manual/image-series/gpt-image-2-5-sunburst-edit) | OpenAI | `openai/gpt-image-2.5-sunburst/edit` |
| [GPT Image 2.6 Flare Text to Image](/api-manual/image-series/gpt-image-2-6-flare-text-to-image) | OpenAI | `openai/gpt-image-2.6-flare/text-to-image` |
| [GPT Image 2.6 Flare Image Edit](/api-manual/image-series/gpt-image-2-6-flare-edit) | OpenAI | `openai/gpt-image-2.6-flare/edit` |
| [GPT Image 2.6 Sunburst Text to Image](/api-manual/image-series/gpt-image-2-6-sunburst-text-to-image) | OpenAI | `openai/gpt-image-2.6-sunburst/text-to-image` |
| [GPT Image 2.6 Sunburst Image Edit](/api-manual/image-series/gpt-image-2-6-sunburst-edit) | OpenAI | `openai/gpt-image-2.6-sunburst/edit` |
| [Nano Banana Text to Image](/api-manual/image-series/google-nano-banana-text-to-image) | Google | `google/nano-banana/text-to-image` |
| [Nano Banana Edit](/api-manual/image-series/google-nano-banana-edit) | Google | `google/nano-banana/edit` |
| [Nano Banana Pro](/api-manual/image-series/nano-banana-2) | Google | `nano-banana-2` `nano-banana-2-edit` |
| [Nano Banana 2](/api-manual/image-series/nano-banana-2-new) | Google | `nano-banana-2-new` `nano-banana-2-new-edit` |
| [Nano Banana 2 Lite Text to Image](/api-manual/image-series/google-nano-banana-2-lite-text-to-image) | Google | `google/nano-banana-2-lite/edit` `google/nano-banana-2-lite/text-to-image` |
| [Nano Banana 2 Lite Edit](/api-manual/image-series/google-nano-banana-2-lite-edit) | Google | `google/nano-banana-2-lite/edit` |
| [Nano Banana 2.1 Text to Image](/api-manual/image-series/nano-banana-2-1-text-to-image) | Google | `google/nano-banana-2.1/text-to-image` |
| [Nano Banana 2.1 Image Edit](/api-manual/image-series/nano-banana-2-1-edit) | Google | `google/nano-banana-2.1/edit` |
| [Flux Schnell Text to Image](/api-manual/image-series/blackforestlabs-flux-schnell) | Flux | `blackforestlabs/flux/schnell` |
| [Flux Dev Text to Image](/api-manual/image-series/blackforestlabs-flux-dev) | Flux | `blackforestlabs/flux/dev` |
| [Flux Dev Image to Image](/api-manual/image-series/blackforestlabs-flux-dev-image-to-image) | Flux | `blackforestlabs/flux/dev/image-to-image` |
| [Flux.2](/api-manual/image-series/flux-2) | Flux | `flux-2-flex` `flux-2-flex-edit` `flux-2-pro` `flux-2-pro-edit` |
| [FLUX.2 Pro Text to Image](/api-manual/image-series/blackforestlabs-flux-2-pro-text-to-image) | Flux | `blackforestlabs/flux-2-pro/text-to-image` |
| [FLUX.2 Pro Edit](/api-manual/image-series/blackforestlabs-flux-2-pro-edit) | Flux | `blackforestlabs/flux-2-pro/edit` |
| [FLUX.2 Flex Text to Image](/api-manual/image-series/blackforestlabs-flux-2-flex-text-to-image) | Flux | `blackforestlabs/flux-2-flex/text-to-image` |
| [FLUX.2 Flex Edit](/api-manual/image-series/blackforestlabs-flux-2-flex-edit) | Flux | `blackforestlabs/flux-2-flex/edit` |
| [Flux Kontext Pro Text to Image](/api-manual/image-series/blackforestlabs-flux-kontext-pro-text-to-image) | Flux | `blackforestlabs/flux-kontext-pro/text-to-image` |
| [Flux Kontext Pro Edit](/api-manual/image-series/blackforestlabs-flux-kontext-pro) | Flux | `blackforestlabs/flux-kontext-pro` |
| [Flux Kontext Max Text to Image](/api-manual/image-series/blackforestlabs-flux-kontext-max-text-to-image) | Flux | `blackforestlabs/flux-kontext-max/text-to-image` |
| [Flux Kontext Max Edit](/api-manual/image-series/blackforestlabs-flux-kontext-max) | Flux | `blackforestlabs/flux-kontext-max` |
| [FLUX 3 Klein Text to Image](/api-manual/image-series/blackforestlabs-flux-3-klein-text-to-image) | Flux | `blackforestlabs/flux-3-klein/text-to-image` |
| [FLUX 3 Klein Edit](/api-manual/image-series/blackforestlabs-flux-3-klein-edit) | Flux | `blackforestlabs/flux-3-klein/edit` |
| [Seedream-4](/api-manual/image-series/seedream-4) | Seedream | `seedream-4` `seedream-4-edit` `seedream-v4` `seedream-v4-edit` |
| [Seedream 4 Text to Image](/api-manual/image-series/bytedance-seedream-v4-text-to-image) | Seedream | `bytedance/seedream/v4/text-to-image` |
| [Seedream 4 Edit](/api-manual/image-series/bytedance-seedream-v4-edit) | Seedream | `bytedance/seedream/v4/edit` |
| [Seedream 4.5 Text to Image](/api-manual/image-series/bytedance-seedream-v4-5-text-to-image) | Seedream | `bytedance/seedream/v4.5/text-to-image` |
| [Seedream 4.5 Edit](/api-manual/image-series/bytedance-seedream-v4-5-edit) | Seedream | `bytedance/seedream/v4.5/edit` |
| [Seedream 5.0 Lite Text to Image](/api-manual/image-series/bytedance-seedream-v5-lite-text-to-image) | Seedream | `bytedance/seedream/v5/lite/text-to-image` |
| [Seedream 5.0 Lite Edit](/api-manual/image-series/bytedance-seedream-v5-lite-edit) | Seedream | `bytedance/seedream/v5/lite/edit` |
| [Seedream 5.0 Pro Text to Image](/api-manual/image-series/seedream-v5-0-pro) | Seedream | `seedream-v5.0-pro` |
| [Seedream 5.0 Pro Image Edit](/api-manual/image-series/seedream-v5-0-pro-edit) | Seedream | `seedream-v5.0-pro/edit` |
| [Seedream 5.0 Flash Text to Image](/api-manual/image-series/seedream-v5-0-flash) | Seedream | `bytedance/seedream-v5.0-flash/text-to-image` |
| [Seedream 5.0 Flash Image Edit](/api-manual/image-series/seedream-v5-0-flash-edit) | Seedream | `bytedance/seedream-v5.0-flash/edit` |
| [Kling O1 Image Edit](/api-manual/image-series/kwaivgi-kling-image-o1) | Kling | `kwaivgi/kling-image-o1` |
| [Kling O3 Image Text to Image](/api-manual/image-series/kwaivgi-kling-image-o3-text-to-image) | Kling | `kwaivgi/kling-image-o3/text-to-image` |
| [Kling O3 Image Edit](/api-manual/image-series/kwaivgi-kling-image-o3-edit) | Kling | `kwaivgi/kling-image-o3/edit` |
| [Qwen Image 2.1 Text to Image](/api-manual/image-series/qwen-image-2-1-text-to-image) | Alibaba | `alibaba/qwen-image-2.1/text-to-image` |
| [Qwen Image 2.1 Image Edit](/api-manual/image-series/qwen-image-2-1-edit) | Alibaba | `alibaba/qwen-image-2.1/edit` |
| [Qwen Image 3.0 Text to Image](/api-manual/image-series/qwen-image-3-0-text-to-image) | Alibaba | `qwen-image-3.0/text-to-image` |
| [Qwen Image 3.0 Image Edit](/api-manual/image-series/qwen-image-3-0-edit) | Alibaba | `qwen-image-3.0/edit` |
| [Qwen Image 3.0 Pro Text to Image](/api-manual/image-series/qwen-image-3-0-pro-text-to-image) | Alibaba | `qwen-image-3.0-pro/text-to-image` |
| [Qwen Image 3.0 Pro Image Edit](/api-manual/image-series/qwen-image-3-0-pro-edit) | Alibaba | `qwen-image-3.0-pro/edit` |
| [Wan 2.7 Image Text to Image](/api-manual/image-series/alibaba-wan-2-7-text-to-image) | Alibaba | `alibaba/wan-2.7/text-to-image` |
| [Wan 2.7 Image Edit](/api-manual/image-series/alibaba-wan-2-7-image-edit) | Alibaba | `alibaba/wan-2.7/image-edit` |
| [Wan 2.7 Pro Text to Image](/api-manual/image-series/alibaba-wan-2-7-text-to-image-pro) | Alibaba | `alibaba/wan-2.7/text-to-image-pro` |
| [Wan 2.7 Pro Image Edit](/api-manual/image-series/alibaba-wan-2-7-image-edit-pro) | Alibaba | `alibaba/wan-2.7/image-edit-pro` |
| [Z-Image Turbo](/api-manual/image-series/alibaba-z-image-turbo) | Alibaba | `alibaba/z-image/turbo` |
| [Grok Imagine Text to Image](/api-manual/image-series/xai-grok-imagine-image-text-to-image) | Grok (xAI) | `xai/grok-imagine-image/text-to-image` |
| [Grok Imagine Edit](/api-manual/image-series/xai-grok-imagine-image-edit) | Grok (xAI) | `xai/grok-imagine-image/edit` |
| [Grok Imagine Image 2.0 Text to Image](/api-manual/image-series/xai-grok-imagine-image-v2-0-text-to-image) | Grok (xAI) | `xai/grok-imagine-image-v2.0/text-to-image` |
| [Grok Imagine Image 2.0 Edit](/api-manual/image-series/xai-grok-imagine-image-v2-0-edit) | Grok (xAI) | `xai/grok-imagine-image-v2.0/edit` |
| [Grok Imagine Image Quality Text to Image](/api-manual/image-series/xai-grok-imagine-image-quality-text-to-image) | Grok (xAI) | `xai/grok-imagine-image-quality/text-to-image` |
| [Grok Imagine Image Quality Edit](/api-manual/image-series/xai-grok-imagine-image-quality-edit) | Grok (xAI) | `xai/grok-imagine-image-quality/edit` |
| [Ideogram 4.5 Text to Image](/api-manual/image-series/ideogram-v4-5-text-to-image) | Ideogram | `ideogram/v4.5/text-to-image` |
| [Ideogram 4.5 Edit](/api-manual/image-series/ideogram-v4-5-edit) | Ideogram | `ideogram/v4.5/edit` |
| [Ideogram Vector 1.0 Text to SVG](/api-manual/image-series/ideogram-vector-1-0-text-to-vector) | Ideogram | `ideogram/vector-1.0/text-to-vector` |
| [Ideogram Vector 1.0 Image to SVG](/api-manual/image-series/ideogram-vector-1-0-image-to-vector) | Ideogram | `ideogram/vector-1.0/image-to-vector` |

## Notes

* Each model page lists the parameters, limits, and modes for its model IDs. Parameters are not shared across models.
* Generated files are valid for 24 hours after creation. Download and store them promptly.
* URL inputs must be publicly accessible HTTP(S) URLs. To use local files, upload them first with [File Management](/api-manual/file-series/upload-url).


# Qwen Image 2.1 Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/qwen-image-2-1-edit

api-manual/image-series/qwen-image-2-1-edit.json POST /api/generate/submit
Edit one image with Qwen Image 2.1 using text instructions and reference images.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `size` (string): Output aspect ratio. auto follows the reference image. Ignored when mask\_url is supplied. Accepted values: `1:1`, `4:3`, `3:4`, `3:2`, `2:3`, `16:9`, `9:16`, `21:9`, `9:21`, `auto`. Default: `auto`.
* `resolution` (string): Output resolution tier. Accepted values: `1K`, `2K`. Default: `1K`.
* `output_format` (string): Output image format. Transparent output requires PNG or WebP. Accepted values: `png`, `jpeg`, `webp`. Default: `png`.
* `enable_safety_checker` (boolean): Enable content safety checking. Default: `true`.
* `background` (string): Background mode. Transparent output requires PNG or WebP. Accepted values: `opaque`, `transparent`. Default: `opaque`.
* `mask_url` (string): Optional local-edit mask. Requires exactly one reference image and an opaque background. White areas change and black areas stay unchanged. Match the reference aspect ratio; the mask is scaled to the reference. size and prompt\_extend are ignored.
* `prompt_extend` (boolean): Enhance the prompt before generation. Ignored when mask\_url is supplied. Default: `true`.
* `seed` (integer): Optional random seed. Omit to use a random seed.

## Usage Guide

Edit one image with Qwen Image 2.1 using text instructions and reference images.

* Set `model` to `alibaba/qwen-image-2.1/edit`.
* Place `model` and optional `callback_url` at the request root; place workflow parameters inside `input`.
* Each request produces one image.
* For prompt-only generation, use [Text to Image](/api-manual/image-series/qwen-image-2-1-text-to-image) with `model=alibaba/qwen-image-2.1/text-to-image`.

## Parameter Details

* `prompt` is required. Supply non-blank text of up to 5,000 characters.
* `image_urls` is required and accepts 1 to 10 reference image URLs. Use publicly accessible HTTP(S) URLs in the same order as the references in your prompt.
* `size` defaults to `auto`. `auto` follows the reference image aspect ratio.
* A transparent background requires `output_format=png` or `webp`.
* `mask_url` requires exactly one reference image and `background=opaque`. White areas change; black areas stay unchanged. Match the reference aspect ratio; the mask is scaled to the reference dimensions.
* With a mask, output follows the reference aspect ratio; `size` and `prompt_extend` are ignored.
* Reference images support JPEG, PNG, and WebP, up to 30 MB and 25 megapixels per image. Media limits are checked during generation.

## Billing

Billing is per output image and depends on `resolution`. Text-to-image and editing use the same rate at each tier.

Credits are deducted on submission and refunded if the task fails.


# Qwen Image 2.1 Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/qwen-image-2-1-text-to-image

api-manual/image-series/qwen-image-2-1-text-to-image.json POST /api/generate/submit
Generate one image with Qwen Image 2.1 using a text prompt.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `size` (string): Output aspect ratio. Accepted values: `1:1`, `4:3`, `3:4`, `3:2`, `2:3`, `16:9`, `9:16`, `21:9`, `9:21`. Default: `1:1`.
* `resolution` (string): Output resolution tier. Accepted values: `1K`, `2K`. Default: `1K`.
* `output_format` (string): Output image format. Transparent output requires PNG or WebP. Accepted values: `png`, `jpeg`, `webp`. Default: `png`.
* `enable_safety_checker` (boolean): Enable content safety checking. Default: `true`.
* `background` (string): Background mode. Transparent output requires PNG or WebP. Accepted values: `opaque`, `transparent`. Default: `opaque`.
* `prompt_extend` (boolean): Enhance the prompt before generation. Default: `true`.
* `seed` (integer): Optional random seed. Omit to use a random seed.

## Usage Guide

Generate one image with Qwen Image 2.1 using a text prompt.

* Set `model` to `alibaba/qwen-image-2.1/text-to-image`.
* Place `model` and optional `callback_url` at the request root; place workflow parameters inside `input`.
* Each request produces one image.
* For reference-image editing, use [Image Edit](/api-manual/image-series/qwen-image-2-1-edit) with `model=alibaba/qwen-image-2.1/edit`.

## Parameter Details

* `prompt` is required. Supply non-blank text of up to 5,000 characters.
* `size` defaults to `1:1`.
* A transparent background requires `output_format=png` or `webp`.

## Billing

Billing is per output image and depends on `resolution`. Text-to-image and editing use the same rate at each tier.

Credits are deducted on submission and refunded if the task fails.


# Qwen Image 3.0 Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/qwen-image-3-0-edit

api-manual/image-series/qwen-image-3-0-edit.json POST /api/generate/submit
Qwen Image 3 standard reference-image editing model with 1K and 2K output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

This model supports reference-image editing for posters, product visuals, typography, and natural-language creative tasks.

## Models

* **qwen-image-3.0/edit** - Standard reference-image editing with 1K and 2K output

## Parameters

* `prompt` is required and supports up to 5,000 characters.
* `size` supports `1:1`, `3:2`, `2:3`, `4:3`, `3:4`, `16:9`, `9:16`, and `21:9`.
* `resolution` supports `1K` and `2K`.
* `output_format` supports `png` and `jpeg`.
* `prompt_extend` controls automatic prompt enhancement.
* `negative_prompt` describes content that should not appear in the output.
* `seed` accepts values from 0 to 2,147,483,647.
* `enable_safety_checker` controls content safety checking.
* `image_urls` is required.
* Billing is per output image, with an additional usage charge for each input image.

## Reference-Image Example

```json theme={null}
{
  "model": "qwen-image-3.0/edit",
  "callback_url": "https://your-domain.com/callback",
  "input": {
    "prompt": "A cinematic scene with natural lighting and coherent motion",
    "resolution": "1K",
    "size": "1:1",
    "image_urls": [
      "https://cdn.vidgo.ai/temp/reference-1.jpg"
    ]
  }
}
```


# Qwen Image 3.0 Pro Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/qwen-image-3-0-pro-edit

api-manual/image-series/qwen-image-3-0-pro-edit.json POST /api/generate/submit
Qwen Image 3 higher-quality reference-image editing model with 1K and 2K output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

This model supports reference-image editing for posters, product visuals, typography, and natural-language creative tasks.

## Models

* **qwen-image-3.0-pro/edit** - Higher-quality reference-image editing with 1K and 2K output

## Parameters

* `prompt` is required and supports up to 5,000 characters.
* `size` supports `1:1`, `3:2`, `2:3`, `4:3`, `3:4`, `16:9`, `9:16`, and `21:9`.
* `resolution` supports `1K` and `2K`.
* `output_format` supports `png` and `jpeg`.
* `prompt_extend` controls automatic prompt enhancement.
* `negative_prompt` describes content that should not appear in the output.
* `seed` accepts values from 0 to 2,147,483,647.
* `enable_safety_checker` controls content safety checking.
* `image_urls` is required.
* Billing is per output image; resolution and the number of input images can affect the current model price.

## Reference-Image Example

```json theme={null}
{
  "model": "qwen-image-3.0-pro/edit",
  "callback_url": "https://your-domain.com/callback",
  "input": {
    "prompt": "A cinematic scene with natural lighting and coherent motion",
    "resolution": "1K",
    "size": "1:1",
    "image_urls": [
      "https://cdn.vidgo.ai/temp/reference-1.jpg"
    ]
  }
}
```


# Qwen Image 3.0 Pro Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/qwen-image-3-0-pro-text-to-image

api-manual/image-series/qwen-image-3-0-pro-text-to-image.json POST /api/generate/submit
Qwen Image 3 higher-quality text-to-image generation model with 1K and 2K output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

This model supports text-to-image generation for posters, product visuals, typography, and natural-language creative tasks.

## Models

* **qwen-image-3.0-pro/text-to-image** - Higher-quality text-to-image generation with 1K and 2K output

## Parameters

* `prompt` is required and supports up to 5,000 characters.
* `size` supports `1:1`, `3:2`, `2:3`, `4:3`, `3:4`, `16:9`, `9:16`, and `21:9`.
* `resolution` supports `1K` and `2K`.
* `output_format` supports `png` and `jpeg`.
* `prompt_extend` controls automatic prompt enhancement.
* `negative_prompt` describes content that should not appear in the output.
* `seed` accepts values from 0 to 2,147,483,647.
* `enable_safety_checker` controls content safety checking.
* Billing is per output image, and resolution can affect the current model price.

## Text-to-Image Example

```json theme={null}
{
  "model": "qwen-image-3.0-pro/text-to-image",
  "callback_url": "https://your-domain.com/callback",
  "input": {
    "prompt": "A cinematic scene with natural lighting and coherent motion",
    "resolution": "1K",
    "size": "1:1"
  }
}
```


# Qwen Image 3.0 Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/qwen-image-3-0-text-to-image

api-manual/image-series/qwen-image-3-0-text-to-image.json POST /api/generate/submit
Qwen Image 3 standard text-to-image generation model with 1K and 2K output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

This model supports text-to-image generation for posters, product visuals, typography, and natural-language creative tasks.

## Models

* **qwen-image-3.0/text-to-image** - Standard text-to-image generation with 1K and 2K output

## Parameters

* `prompt` is required and supports up to 5,000 characters.
* `size` supports `1:1`, `3:2`, `2:3`, `4:3`, `3:4`, `16:9`, `9:16`, and `21:9`.
* `resolution` supports `1K` and `2K`.
* `output_format` supports `png` and `jpeg`.
* `prompt_extend` controls automatic prompt enhancement.
* `negative_prompt` describes content that should not appear in the output.
* `seed` accepts values from 0 to 2,147,483,647.
* `enable_safety_checker` controls content safety checking.
* Billing is per output image using the current model price.

## Text-to-Image Example

```json theme={null}
{
  "model": "qwen-image-3.0/text-to-image",
  "callback_url": "https://your-domain.com/callback",
  "input": {
    "prompt": "A cinematic scene with natural lighting and coherent motion",
    "resolution": "1K",
    "size": "1:1"
  }
}
```


# Seedream-4
Source: https://docs.vidgo.ai/api-manual/image-series/seedream-4

api-manual/image-series/seedream-4.json POST /api/generate/submit
Image generation model powered by ByteDance Seedream-4 with support for text-to-image and image editing

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Seedream-4 is powered by ByteDance Seedream-4. It supports both text-to-image generation and image editing through the same unified submit endpoint.

## Models

* **seedream-4** - Text-to-image generation, optionally with reference images
* **seedream-4-edit** - Image editing mode. `image_urls` is required
* **seedream-v4** / **seedream-v4-edit** - Legacy aliases of `seedream-4` / `seedream-4-edit` with the same parameters and pricing

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [bytedance/seedream/v4/text-to-image](/api-manual/image-series/bytedance-seedream-v4-text-to-image)
* [bytedance/seedream/v4/edit](/api-manual/image-series/bytedance-seedream-v4-edit)

## Notes

* `size` maps to aspect ratio and supports `1:1`, `3:4`, `4:3`, `16:9`, `9:16`, `3:2`, `2:3`, `21:9`
* `resolution` supports `1K`, `2K`, `4K`, with default `2K`
* `n` supports `1-15`
* The total of `image_urls` plus `n` must not exceed `15`
* Credits are pre-deducted as `base credits * n`. If fewer images are returned than requested, the unused portion is refunded automatically


# Seedream 5.0 Flash Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/seedream-v5-0-flash

api-manual/image-series/seedream-v5-0-flash.json POST /api/generate/submit
Generate one image with Seedream 5.0 Flash using a text prompt.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `size` (string): Output aspect ratio. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `2:3`, `3:2`, `21:9`. Default: `1:1`.
* `resolution` (string): Output resolution tier. Accepted values: `1K`, `1.5K`, `2K`. Default: `1K`.
* `output_format` (string): Output image format. Accepted values: `jpeg`, `png`. If omitted, the model selects the format.
* `enable_safety_checker` (boolean): Enable content safety checking. Default: `true`.

## Usage Guide

Generate one image with Seedream 5.0 Flash using a text prompt.

* Set `model` to `bytedance/seedream-v5.0-flash/text-to-image`.
* Place `model` and optional `callback_url` at the request root; place workflow parameters inside `input`.
* Each request produces one image.
* For reference-image editing, use [Image Edit](/api-manual/image-series/seedream-v5-0-flash-edit) with `model=bytedance/seedream-v5.0-flash/edit`.

## Parameter Details

* `prompt` is required. Supply non-blank text of up to 5,000 characters.

## Billing

Billing is per output image. Text-to-image and editing use the same rate across the supported resolution tiers.

Credits are deducted on submission and refunded if the task fails.


# Seedream 5.0 Flash Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/seedream-v5-0-flash-edit

api-manual/image-series/seedream-v5-0-flash-edit.json POST /api/generate/submit
Edit one image with Seedream 5.0 Flash using text instructions and reference images.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `size` (string): Output aspect ratio. Accepted values: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `2:3`, `3:2`, `21:9`. Default: `1:1`.
* `resolution` (string): Output resolution tier. Accepted values: `1K`, `1.5K`, `2K`. Default: `1K`.
* `output_format` (string): Output image format. Accepted values: `jpeg`, `png`. If omitted, the model selects the format.
* `enable_safety_checker` (boolean): Enable content safety checking. Default: `true`.

## Usage Guide

Edit one image with Seedream 5.0 Flash using text instructions and reference images.

* Set `model` to `bytedance/seedream-v5.0-flash/edit`.
* Place `model` and optional `callback_url` at the request root; place workflow parameters inside `input`.
* Each request produces one image.
* For prompt-only generation, use [Text to Image](/api-manual/image-series/seedream-v5-0-flash) with `model=bytedance/seedream-v5.0-flash/text-to-image`.

## Parameter Details

* `prompt` is required. Supply non-blank text of up to 5,000 characters.
* `image_urls` is required and accepts 1 to 10 reference image URLs. Use publicly accessible HTTP(S) URLs in the same order as the references in your prompt.

## Billing

Billing is per output image. Text-to-image and editing use the same rate across the supported resolution tiers.

Credits are deducted on submission and refunded if the task fails.


# Seedream 5.0 Pro Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/seedream-v5-0-pro

api-manual/image-series/seedream-v5-0-pro.json POST /api/generate/submit
Flagship Seedream text-to-image model with precise prompt following, layout control, and multilingual typography

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Seedream 5.0 Pro is designed for complex prompts, structured layouts, multilingual typography, and product visuals.

## Models

* **seedream-v5.0-pro** - Flagship text-to-image generation

## Size Parameter

`input.size` controls the output aspect ratio and defaults to `1:1`:

* Ratio presets: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `2:3`, `3:2`, `21:9`

## Resolution Parameter

`input.resolution` controls output resolution and defaults to `1K`:

* Resolution presets: `1K`, `2K`

## Pricing

Each request generates one image and uses the current Seedream 5.0 Pro model price.

## How to Call

Use the unified submit endpoint: `POST /api/generate/submit`

```json theme={null}
{
  "model": "seedream-v5.0-pro",
  "callback_url": "https://your-domain.com/callback",
  "input": {
    "prompt": "A cinematic scene with natural lighting and coherent motion",
    "resolution": "1K",
    "size": "1:1"
  }
}
```


# Seedream 5.0 Pro Image Edit
Source: https://docs.vidgo.ai/api-manual/image-series/seedream-v5-0-pro-edit

api-manual/image-series/seedream-v5-0-pro-edit.json POST /api/generate/submit
Flagship Seedream image editing model with precise prompt following and multi-image reference control

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Seedream 5.0 Pro Edit applies controlled changes using up to ten reference images while preserving requested visual details.

## Models

* **seedream-v5.0-pro/edit** - Image editing with multi-image reference support

## Size Parameter

`input.size` controls the output aspect ratio and defaults to `1:1`:

* Ratio presets: `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `2:3`, `3:2`, `21:9`

## Resolution Parameter

`input.resolution` controls output resolution and defaults to `1K`:

* Resolution presets: `1K`, `2K`

## Pricing

Each request generates one image and uses the current Seedream 5.0 Pro model price.

## How to Call

Use the unified submit endpoint: `POST /api/generate/submit`

```json theme={null}
{
  "model": "seedream-v5.0-pro/edit",
  "callback_url": "https://your-domain.com/callback",
  "input": {
    "prompt": "A cinematic scene with natural lighting and coherent motion",
    "resolution": "1K",
    "size": "1:1",
    "image_urls": [
      "https://cdn.vidgo.ai/temp/reference-1.jpg"
    ]
  }
}
```


# Grok Imagine Edit
Source: https://docs.vidgo.ai/api-manual/image-series/xai-grok-imagine-image-edit

api-manual/image-series/xai-grok-imagine-image-edit.json POST /api/generate/submit
Reimagine an existing image with Grok Imagine, using text to guide changes to the subject, setting, or visual style.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Reimagine an existing image with Grok Imagine, using text to guide changes to the subject, setting, or visual style.

## Models

* **xai/grok-imagine-image/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts exactly 1 image in `image_urls`.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `9:16`, `16:9`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below. Supply publicly accessible HTTP(S) URLs for reference images.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-5,000 characters**.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Use JPEG, PNG, or WebP, up to 10 MB.


# Grok Imagine Image Quality Edit
Source: https://docs.vidgo.ai/api-manual/image-series/xai-grok-imagine-image-quality-edit

api-manual/image-series/xai-grok-imagine-image-quality-edit.json POST /api/generate/submit
Develop new image variations with Grok Imagine Image Quality, combining text instructions and visual references in output up to 2K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Develop new image variations with Grok Imagine Image Quality, combining text instructions and visual references in output up to 2K.

## Models

* **xai/grok-imagine-image-quality/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-3 images in `image_urls`.
* Choose from `1K`, `2K` output resolutions.
* Control the number of output images with `n`.

## Output Options

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `9:16`, `16:9`. Default: `1:1`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`. Default: `1K`.

### Image Count

* `n`: Optional. Number of images to generate. Integer, at least `1`. Default: `1`. No maximum is specified; fractional numbers, strings, and booleans are rejected.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes. Minimum length: **1 character**.
* The value must be a string and must not be blank. No maximum prompt length is specified.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-3 items. Each item must be an HTTP(S) URL.

## Billing Factors

Billing depends on output resolution (`resolution`) and output image count (`n`). Reference image count (`image_urls`) adds a separate input charge once per request; this charge is not multiplied by `n`.

Credits are deducted on submission and refunded if the task fails.


# Grok Imagine Image Quality Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/xai-grok-imagine-image-quality-text-to-image

api-manual/image-series/xai-grok-imagine-image-quality-text-to-image.json POST /api/generate/submit
Create image variations in 1K or 2K with Grok Imagine Image Quality, with framing options for portraits, landscapes, and square designs.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create image variations in 1K or 2K with Grok Imagine Image Quality, with framing options for portraits, landscapes, and square designs.

## Models

* **xai/grok-imagine-image-quality/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Choose from `1K`, `2K` output resolutions.
* Control the number of output images with `n`.

## Output Options

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `9:16`, `16:9`. Default: `1:1`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`. Default: `1K`.

### Image Count

* `n`: Optional. Number of images to generate. Integer, at least `1`. Default: `1`. No maximum is specified; fractional numbers, strings, and booleans are rejected.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the desired image. Minimum length: **1 character**.
* The value must be a string and must not be blank. No maximum prompt length is specified.

## Billing Factors

Billing depends on output resolution (`resolution`) and output image count (`n`).

Credits are deducted on submission and refunded if the task fails.


# Grok Imagine Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/xai-grok-imagine-image-text-to-image

api-manual/image-series/xai-grok-imagine-image-text-to-image.json POST /api/generate/submit
Turn text prompts into images with Grok Imagine, exploring visual ideas in square, portrait, and landscape formats.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn text prompts into images with Grok Imagine, exploring visual ideas in square, portrait, and landscape formats.

## Models

* **xai/grok-imagine-image/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.

## Output Options

### Image Size

* `size`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `9:16`, `16:9`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-5,000 characters**.


# Grok Imagine Image 2.0 Edit
Source: https://docs.vidgo.ai/api-manual/image-series/xai-grok-imagine-image-v2-0-edit

api-manual/image-series/xai-grok-imagine-image-v2-0-edit.json POST /api/generate/submit
Refine images with Grok Imagine Image 2.0, combining reference images and precise instructions to preserve subjects while changing visual details.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Refine images with Grok Imagine Image 2.0, combining reference images and precise instructions to preserve subjects while changing visual details.

## Models

* **xai/grok-imagine-image-v2.0/edit** - Edit images using a prompt and image references.

## Features

* Edit images using a prompt and image references.
* Accepts 1-3 images in `image_urls`.
* Choose from `1K`, `2K` output resolutions.
* Control the number of output images with `n`.

## Output Options

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `9:16`, `16:9`. Default: `1:1`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`. Default: `1K`.

### Quality

* `quality`: Optional. Accepted values: `low`, `medium`. Default: `medium`.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `4`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes. Length: **1-8,000 characters**.
* The value must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-3 items. Each item must be an HTTP(S) URL.

## Billing Factors

Billing depends on output quality (`quality`), output resolution (`resolution`), and output image count (`n`). Input image count (`image_urls`) adds a separate charge once per request; this charge is not multiplied by `n`.

Credits are deducted on submission and refunded if the task fails.


# Grok Imagine Image 2.0 Text to Image
Source: https://docs.vidgo.ai/api-manual/image-series/xai-grok-imagine-image-v2-0-text-to-image

api-manual/image-series/xai-grok-imagine-image-v2-0-text-to-image.json POST /api/generate/submit
Create detailed images with accurate text, structured layouts, and close prompt following, at resolutions up to 2K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create detailed images with accurate text, structured layouts, and close prompt following, at resolutions up to 2K.

## Models

* **xai/grok-imagine-image-v2.0/text-to-image** - Generate images from text prompts.

## Features

* Generate images from text prompts.
* Choose from `1K`, `2K` output resolutions.
* Control the number of output images with `n`.

## Output Options

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `9:16`, `16:9`. Default: `1:1`.

### Resolution

* `resolution`: Optional. Accepted values: `1K`, `2K`. Default: `1K`.

### Quality

* `quality`: Optional. Accepted values: `low`, `medium`. Default: `medium`.

### Image Count

* `n`: Optional. Number of images to generate. Integer from `1` to `4`. Default: `1`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the desired image. Length: **1-8,000 characters**.
* The value must not be blank when supplied.

## Billing Factors

Billing depends on output quality (`quality`), output resolution (`resolution`), and output image count (`n`).

Credits are deducted on submission and refunded if the task fails.


# Music Webhook
Source: https://docs.vidgo.ai/api-manual/music-series/music-webhook

Receive automatic callback notifications when music generation tasks complete

Instead of polling the [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint, you can provide a `callback_url` when submitting music generation tasks to receive automatic notifications when generation completes.

## How It Works

1. **Submit with callback URL**: Include a `callback_url` parameter in your music generation request
2. **Task processing**: Vidgo API processes your music generation task
3. **Receive notification**: When the task status becomes `finished` or `failed`, Vidgo API sends a POST request to your callback URL
4. **Process result**: Your server receives the complete task data including generated files

## Callback Request

When a task completes, Vidgo API will send a POST request to your callback URL:

### Request Headers

```
Content-Type: application/json
```

### Request Body Structure

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [...],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

> [!NOTE]
>
> The `files` array content varies depending on which model created the task. See [Response Fields](#response-fields-files-data) below for details.

## Response Fields (files data)

The `files` array content varies depending on which model created the task.

| Model | Field | Type | Description |
| - | - | - | - |
| `generate-music`, `extend-music`, `upload-and-cover-audio`, `upload-and-extend-audio`, `add-instrumental`, `add-vocals`, `replace-section` | `audio_id` | string | Unique identifier for the audio file |
| ↑ | `audio_url` | string | Direct download URL for the audio file |
| ↑ | `image_url` | string | Cover image URL |
| ↑ | `title` | string | Track title |
| ↑ | `tags` | string | Style tags |
| ↑ | `duration` | number | Audio duration in seconds |
| ↑ | `prompt` | string | Generation prompt used |
| `get-timestamped-lyrics` | `timestampe_lyrics` | string | Lyrics with timestamps |
| `generate-lyrics` | `title` | string | Lyrics title |
| ↑ | `text` | string | Complete lyrics content |
| `boost-music-style` | `style` | string | Enhanced music style |
| `convert-to-wav` | `wav_url` | string | URL to converted WAV file |
| `separate-vocals` | `separate_vocals` | string (JSON) | JSON string containing separated audio URLs. [See details](#separate-vocals) |
| `upload-and-separate-vocals` | `vocal_removal` | string (JSON) | JSON string containing separated stem URLs. [See details](#upload-and-separate-vocals) |
| `stem-split` | `stem_split` | string (JSON) | JSON string containing full stem separation URLs. [See details](#stem-split) |
| `generate-music-cover` | `generate_cover` | string (JSON) | JSON string containing generated cover images. [See details](#generate-music-cover) |
| `generate-persona` | `persona_id` | string | Generated persona identifier |
| `create-music-video` | `video_url` | string | Generated video URL |

### JSON Field Details

#### separate-vocals

`separate_vocals` fields:

* `vocal_url` - URL to the extracted vocal track
* `instrumental_url` - URL to the instrumental track

#### upload-and-separate-vocals

`vocal_removal` fields:

* `bass` - URL to the bass track
* `drums` - URL to the drums track
* `piano` - URL to the piano track
* `guitar` - URL to the guitar track
* `vocals` - URL to the vocals track
* `other` - URL to other audio elements

#### stem-split

`stem_split` fields:

* `backing_vocals_url` - URL to backing vocals
* `bass_url` - URL to bass track
* `brass_url` - URL to brass instruments
* `drums_url` - URL to drums track
* `fx_url` - URL to sound effects
* `guitar_url` - URL to guitar track
* `keyboard_url` - URL to keyboard track
* `percussion_url` - URL to percussion track
* `strings_url` - URL to strings track
* `synth_url` - URL to synthesizer track
* `vocal_url` - URL to main vocals
* `woodwinds_url` - URL to woodwinds track

#### generate-music-cover

`generate_cover` is a JSON array containing cover image objects:

* `file_url` - URL to the generated cover image
* `file_type` - File type (e.g., "image")

## Callback Examples

### Music Generation (generate-music, extend-music, etc.)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "audio_id": "62e4542a-73be-44e1-b397-7716ee7505c6",
        "audio_url": "https://file.goseaapp.com/audio/8FDN1I7M7Q68DDG8/audio_62e4542a.mp3",
        "image_url": "https://file.goseaapp.com/audio/8FDN1I7M7Q68DDG8/cover_62e4542a.jpeg",
        "title": "Peaceful Piano Meditation",
        "tags": "Classical",
        "duration": 240.0,
        "prompt": ""
      },
      {
        "audio_id": "8e755734-a840-4c13-beac-b9f1044979ca",
        "audio_url": "https://file.goseaapp.com/audio/8FDN1I7M7Q68DDG8/audio_8e755734.mp3",
        "image_url": "https://file.goseaapp.com/audio/8FDN1I7M7Q68DDG8/cover_8e755734.jpeg",
        "title": "Peaceful Piano Meditation",
        "tags": "Classical",
        "duration": 130.0,
        "prompt": ""
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Timestamped Lyrics (get-timestamped-lyrics)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "timestampe_lyrics": "[00:00.00] First line of lyrics\n[00:05.23] Second line of lyrics"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Generated Lyrics (generate-lyrics)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "title": "Peaceful Piano Meditation",
        "text": "Peaceful Piano Meditation lyrics content..."
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Music Style Enhancement (boost-music-style)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "style": "Enhanced music style result"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### WAV Conversion (convert-to-wav)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "wav_url": "https://file.goseaapp.com/audio/8FDN1I7M7Q68DDG8/output.wav"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Vocal Separation (separate-vocals)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "separate_vocals": "{\"vocal_url\": \"https://file.goseaapp.com/audio/vocals.mp3\", \"instrumental_url\": \"https://file.goseaapp.com/audio/instrumental.mp3\"}"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Upload & Separate Vocals (upload-and-separate-vocals)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "vocal_removal": "{\"bass\": \"https://file.goseaapp.com/audio/bass.mp3\", \"drums\": \"https://file.goseaapp.com/audio/drums.mp3\", \"piano\": \"https://file.goseaapp.com/audio/piano.mp3\", \"guitar\": \"https://file.goseaapp.com/audio/guitar.mp3\", \"vocals\": \"https://file.goseaapp.com/audio/vocals.mp3\", \"other\": \"https://file.goseaapp.com/audio/other.mp3\"}"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Stem Split (stem-split)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "stem_split": "{\"backing_vocals_url\": \"\", \"bass_url\": \"\", \"brass_url\": \"\", \"drums_url\": \"\", \"fx_url\": \"\", \"guitar_url\": \"\", \"keyboard_url\": \"\", \"percussion_url\": \"\", \"strings_url\": \"\", \"synth_url\": \"\", \"vocal_url\": \"\", \"woodwinds_url\": \"\"}"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Music Cover Image (generate-music-cover)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "generate_cover": "[{\"file_url\": \"https://file.goseaapp.com/audio/cover_1.png\", \"file_type\": \"image\"}, {\"file_url\": \"https://file.goseaapp.com/audio/cover_2.png\", \"file_type\": \"image\"}]"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Generated Persona (generate-persona)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "persona_id": "persona_abc123"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Music Video (create-music-video)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "finished",
    "files": [
      {
        "video_url": "https://file.goseaapp.com/video/8FDN1I7M7Q68DDG8/music_video.mp4"
      }
    ],
    "created_time": "2025-11-25T08:50:13",
    "error_message": null
  }
}
```

### Failed Task

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "8FDN1I7M7Q68DDG8",
    "status": "failed",
    "files": [],
    "created_time": "2025-11-25T08:50:13",
    "error_message": "The prompt violates our content policy"
  }
}
```

## Requirements

Your callback endpoint must meet the following requirements:

* **HTTPS only**: Must use HTTPS protocol (HTTP not supported)
* **Maximum URL length**: 2048 characters
* **Response timeout**: Must respond within 10 seconds
* **Success response**: Should return HTTP 200-299 status code
* **No internal IPs**: Cannot use private/internal IP addresses (e.g., 192.168.x.x, 10.x.x.x)
* **Public accessibility**: Must be publicly accessible from the internet

## Retry Policy

If your callback endpoint fails to respond or returns an error:

* **Retry attempts**: 3 automatic retries
* **Retry delays**: After 1 second, 2 seconds, and 4 seconds
* **Final failure**: After 3 failed attempts, no further retries are made

> [!WARNING]
>
> If all retry attempts fail, you can still retrieve the results by polling the [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Security Best Practices

> [!TIP]
>
> **Verify requests**: Consider adding a signature verification mechanism to ensure requests are from Vidgo API

> [!TIP]
>
> **Idempotency**: Design your webhook handler to be idempotent in case of duplicate deliveries

> [!TIP]
>
> **Async processing**: Process the callback asynchronously and return 200 quickly to avoid timeouts

> [!TIP]
>
> **Logging**: Log all webhook requests for debugging and monitoring


# Query Music Detail
Source: https://docs.vidgo.ai/api-manual/music-series/query-music-detail

api-manual/music-series/query-music-detail.json GET /api/generate/detail/music
Query detailed results of music generation tasks

## Usage Guide

* All POST submission requests in the Music Series can query their results through this endpoint. Different models will return different [response fields](#response-fields-files-data).

## Parameter Details

* `task_id` (required): The unique task identifier returned from the music generation submission endpoint

## Usage Notes

* Poll this endpoint to check task completion status
* The `files` array will be empty until the task status is `finished`
* When a task fails, check the `error_message` field for details
* Multiple audio files may be returned for a single task as variations are generated

## Response Fields (files data)

The `files` array content varies depending on which model created the task.

| Model | Field | Type | Description |
| - | - | - | - |
| `generate-music`, `extend-music`, `upload-and-cover-audio`, `upload-and-extend-audio`, `add-instrumental`, `add-vocals`, `replace-section` | `audio_id` | string | Unique identifier for the audio file |
| ↑ | `audio_url` | string | Direct download URL for the audio file |
| ↑ | `image_url` | string | Cover image URL |
| ↑ | `title` | string | Track title |
| ↑ | `tags` | string | Style tags |
| ↑ | `duration` | number | Audio duration in seconds |
| ↑ | `prompt` | string | Generation prompt used |
| `get-timestamped-lyrics` | `timestampe_lyrics` | string | Lyrics with timestamps |
| `generate-lyrics` | `title` | string | Lyrics title |
| ↑ | `text` | string | Complete lyrics content |
| `boost-music-style` | `style` | string | Enhanced music style |
| `convert-to-wav` | `wav_url` | string | URL to converted WAV file |
| `separate-vocals` | `separate_vocals` | string (JSON) | JSON string containing separated audio URLs. [See details](#separate-vocals) |
| `upload-and-separate-vocals` | `vocal_removal` | string (JSON) | JSON string containing separated stem URLs. [See details](#upload-and-separate-vocals) |
| `stem-split` | `stem_split` | string (JSON) | JSON string containing full stem separation URLs. [See details](#stem-split) |
| `generate-music-cover` | `generate_cover` | string (JSON) | JSON string containing generated cover images. [See details](#generate-music-cover) |
| `generate-persona` | `persona_id` | string | Generated persona identifier |
| `create-music-video` | `video_url` | string | Generated video URL |

### JSON Field Details

#### separate-vocals

`separate_vocals` fields:

* `vocal_url` - URL to the extracted vocal track
* `instrumental_url` - URL to the instrumental track

#### upload-and-separate-vocals

`vocal_removal` fields:

* `bass` - URL to the bass track
* `drums` - URL to the drums track
* `piano` - URL to the piano track
* `guitar` - URL to the guitar track
* `vocals` - URL to the vocals track
* `other` - URL to other audio elements

#### stem-split

`stem_split` fields:

* `backing_vocals_url` - URL to backing vocals
* `bass_url` - URL to bass track
* `brass_url` - URL to brass instruments
* `drums_url` - URL to drums track
* `fx_url` - URL to sound effects
* `guitar_url` - URL to guitar track
* `keyboard_url` - URL to keyboard track
* `percussion_url` - URL to percussion track
* `strings_url` - URL to strings track
* `synth_url` - URL to synthesizer track
* `vocal_url` - URL to main vocals
* `woodwinds_url` - URL to woodwinds track

#### generate-music-cover

`generate_cover` is a JSON array containing cover image objects:

* `file_url` - URL to the generated cover image
* `file_type` - File type (e.g., "image")


# Quick Start
Source: https://docs.vidgo.ai/api-manual/quickstart

Make your first Vidgo API call: create an API key, submit an image, video or music generation task, and poll the task status for the result.

## Welcome to Vidgo API

Vidgo API provides powerful image and video generation APIs. This guide will help you make your first API call in just a few minutes.

## Step 1: Get Your API Key

1. Visit the [Vidgo API Console](https://vidgo.ai/apis/dashboard/api-key)
2. Sign in or create an account
3. Generate a new API key
4. **Important**: Copy and securely store your API key - it will only be displayed once

## Step 2: Generate Your First Image

All API requests require your API key in the `Authorization` header. Here's a complete example:

```python theme={null}
import requests

url = "https://api.vidgo.ai/api/generate/submit"
headers = {
    "Authorization": "Bearer VIDGO_API_KEY",
    "Content-Type": "application/json"
}

payload = {
    "model": "gpt-4o-image",
    "callback_url": "https://your-domain.com/callback",
    "input": {
        "prompt": "A serene mountain landscape at sunset with vibrant colors",
        "size": "1:1",
        "n": 1
    }
}

response = requests.post(url, json=payload, headers=headers)
result = response.json()
task_id = result["data"]["task_id"]
print(f"Task ID: {task_id}")
```

```javascript theme={null}
const response = await fetch('https://api.vidgo.ai/api/generate/submit', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer VIDGO_API_KEY',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    model: 'gpt-4o-image',
    callback_url: 'https://your-domain.com/callback',
    input: {
      prompt: 'A serene mountain landscape at sunset with vibrant colors',
      size: '1:1',
      n: 1
    }
  })
});

const result = await response.json();
const taskId = result.data.task_id;
console.log(`Task ID: ${taskId}`);
```

```bash theme={null}
curl -X POST https://api.vidgo.ai/api/generate/submit \
  -H "Authorization: Bearer $VIDGO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-image",
    "callback_url": "https://your-domain.com/callback",
    "input": {
      "prompt": "A serene mountain landscape at sunset with vibrant colors",
      "size": "1:1",
      "n": 1
    }
  }'
```

## Step 3: Check Task Status

Vidgo API uses asynchronous processing. Poll the status endpoint to get your results:

```python theme={null}
import requests
import time

task_id = "task-unified-1757165031-uyujaw3d"
url = f"https://api.vidgo.ai/api/generate/status/{task_id}"
headers = {
    "Authorization": "Bearer VIDGO_API_KEY"
}

# Poll until complete
while True:
    response = requests.get(url, headers=headers)
    result = response.json()
    task = result["data"]

    status = task['status']
    progress = task.get('progress', 0)

    print(f"Status: {status}, Progress: {progress}%")

    if status == 'finished':
        print(f"Image URL: {task['files'][0]['file_url']}")
        break
    elif status == 'failed':
        print(f"Error: {task['error_message']}")
        break

    time.sleep(2)  # Wait 2 seconds before next check
```

```javascript theme={null}
const taskId = 'task-unified-1757165031-uyujaw3d';

async function checkStatus() {
  while (true) {
    const response = await fetch(
      `https://api.vidgo.ai/api/generate/status/${taskId}`,
      {
        headers: {
          'Authorization': 'Bearer VIDGO_API_KEY'
        }
      }
    );

    const result = await response.json();
    const task = result.data;
    const { status, progress, files, error_message } = task;

    console.log(`Status: ${status}, Progress: ${progress}%`);

    if (status === 'finished') {
      console.log('Image URL:', files[0].file_url);
      break;
    } else if (status === 'failed') {
      console.log('Error:', error_message);
      break;
    }

    await new Promise(resolve => setTimeout(resolve, 2000));
  }
}

checkStatus();
```

```bash theme={null}
curl -X GET https://api.vidgo.ai/api/generate/status/task-unified-1757165031-uyujaw3d \
  -H "Authorization: Bearer $VIDGO_API_KEY"
```

**Example Response:**

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "task-unified-1757165031-uyujaw3d",
    "status": "finished",
    "progress": 100,
    "files": [
      {
        "file_url": "https://file.goseaapp.com/images/image-abc123.jpg",
        "file_type": "image"
      }
    ],
    "created_time": "2025-11-12T10:30:00",
    "error_message": null
  }
}
```

> [!NOTE]
>
> **Content Validity**: Generated files are valid for 24 hours. Please download and save them promptly.

## Step 4: Try Video Generation

Generate videos using the same pattern. Just change the model and input parameters:

```python theme={null}
payload = {
    "model": "kwaivgi/kling-v3.0-std/text-to-video",
    "callback_url": "https://your-domain.com/callback",
    "input": {
        "prompt": "A time-lapse of a bustling city street transitioning from day to night",
        "duration": 5,
        "aspect_ratio": "16:9",
        "sound": True
    }
}
```

## Step 5: Try Music Generation

Generate music with customizable styles and vocals:

```python theme={null}
payload = {
    "model": "generate-music",
    "callback_url": "https://your-domain.com/callback",
    "input": {
        "prompt": "A calm and relaxing piano track with soft melodies",
        "style": "Classical",
        "title": "Peaceful Piano Meditation",
        "custom_mode": True,
        "instrumental": True,
        "mv": "V5"
    }
}
```

## Support

Need help? We're here to assist:

* **Email**: [support@vidgo.ai](mailto:support@vidgo.ai)
* **Console**: [vidgo.ai](https://vidgo.ai/apis)


# Get Task Status
Source: https://docs.vidgo.ai/api-manual/task-management/status

openapi.json GET /api/generate/status/{task_id}
Query task execution status and retrieve generation results

Query image, video, audio, and 3D generation tasks and retrieve all returned files when complete.

## Endpoint

```http theme={null}
GET https://api.vidgo.ai/api/generate/status/{task_id}
```

## Authentication

This endpoint currently retrieves tasks by `task_id` without requiring an Authorization header or checking task ownership. The examples include a Bearer header so they can share the same client configuration as authenticated submit requests.

```bash theme={null}
Authorization: Bearer VIDGO_API_KEY
```

> [card] [Get API Key](https://vidgo.ai/apis/dashboard/api-key)
>
> Generate your API key from the console

## Parameters

### Path Parameters

| Parameter | Type | Required | Description |
| - | - | - | - |
| `task_id` | string | Yes | Unique task identifier returned from the submit endpoint |

## Response

### Success Response (200)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "task-unified-1757165031-uyujaw3d",
    "status": "finished",
    "files": [
      {
        "file_url": "https://file.goseaapp.com/images/image-abc123.jpg",
        "file_type": "image"
      }
    ],
    "created_time": "2025-11-12T10:30:00",
    "progress": 100,
    "error_message": null
  }
}
```

### Response Fields

| Field | Type | Description |
| - | - | - |
| `code` | integer | HTTP status code (200 for success) |
| `data.task_id` | string | Unique task identifier |
| `data.status` | string | Task status: `not_started`, `running`, `finished`, `failed` |
| `data.files` | array | Generated media files (when status is `finished`) |
| `data.files[].file_url` | string | Direct URL to generated file |
| `data.files[].file_type` | string | Artifact type: `image`, `video`, `audio`, `lyrics`, `3d`, or `other` |
| `data.files[].watermark_url` | string | Optional watermarked variant |
| `data.files[].label` | string | Optional artifact role, such as model, preview, or timestamps |
| `data.files[].format` | string | Optional file format |
| `data.files[].content_type` | string | Optional MIME type |
| `data.files[].file_name` | string | Optional file name |
| `data.files[].file_size` | integer | Optional size in bytes |
| `data.files[].metadata` | object | Optional additional artifact metadata |
| `data.created_time` | string | ISO 8601 timestamp of task creation |
| `data.progress` | integer or null | Completion percentage when supplied by the service; some tasks may remain at 0 until completion |
| `data.error_message` | string or null | Error description when available; null when there is no error |

### Status Values

| Status | Description |
| - | - |
| `not_started` | Task is queued, waiting to be processed |
| `running` | Task is currently being generated |
| `finished` | Task completed successfully, results available in `files` |
| `failed` | Task failed, error details in `error_message` |

## Response Examples by Status

### Not Started

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "task-unified-1757165031-uyujaw3d",
    "status": "not_started",
    "files": [],
    "created_time": "2025-11-12T10:30:00",
    "progress": 0,
    "error_message": null
  }
}
```

### Running (Image)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "task-unified-1757165031-uyujaw3d",
    "status": "running",
    "files": [],
    "created_time": "2025-11-12T10:30:00",
    "progress": 45,
    "error_message": null
  }
}
```

### Finished (Image)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "task-unified-1757165031-uyujaw3d",
    "status": "finished",
    "files": [
      {
        "file_url": "https://file.goseaapp.com/images/image-abc123.jpg",
        "file_type": "image"
      }
    ],
    "created_time": "2025-11-12T10:30:00",
    "progress": 100,
    "error_message": null
  }
}
```

### Finished (Video)

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "task-unified-1757165211-xyz789",
    "status": "finished",
    "files": [
      {
        "file_url": "https://file.goseaapp.com/videos/video-xyz789.mp4",
        "file_type": "video"
      }
    ],
    "created_time": "2025-11-12T10:30:00",
    "progress": 100,
    "error_message": null
  }
}
```

### Failed

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "task-unified-1757165031-uyujaw3d",
    "status": "failed",
    "files": [],
    "created_time": "2025-11-12T10:30:00",
    "progress": 0,
    "error_message": "The prompt violates our content policy"
  }
}
```

## Code Examples

### cURL

```bash theme={null}
curl -X GET https://api.vidgo.ai/api/generate/status/task-unified-1757165031-uyujaw3d \
  -H "Authorization: Bearer VIDGO_API_KEY"
```

### Python

```python theme={null}
import requests

API_KEY = "your-api-key-here"
task_id = "task-unified-1757165031-uyujaw3d"

response = requests.get(
    f"https://api.vidgo.ai/api/generate/status/{task_id}",
    headers={"Authorization": f"Bearer {API_KEY}"}
)

result = response.json()
data = result["data"]

print(f"Status: {data['status']}")
print(f"Progress: {data['progress']}%")

if data["status"] == "finished":
    for file in data["files"]:
        print(f"Generated file: {file['file_url']}")
elif data["status"] == "failed":
    print(f"Error: {data['error_message']}")
```

### JavaScript / Node.js

```javascript theme={null}
const API_KEY = 'your-api-key-here';
const taskId = 'task-unified-1757165031-uyujaw3d';

const response = await fetch(
  `https://api.vidgo.ai/api/generate/status/${taskId}`,
  {
    headers: {
      'Authorization': `Bearer ${API_KEY}`
    }
  }
);

const result = await response.json();
const data = result.data;

console.log(`Status: ${data.status}`);
console.log(`Progress: ${data.progress}%`);

if (data.status === 'finished') {
  data.files.forEach(file => {
    console.log(`Generated file: ${file.file_url}`);
  });
} else if (data.status === 'failed') {
  console.log(`Error: ${data.error_message}`);
}
```

### Go

```go theme={null}
package main

import (
    "encoding/json"
    "fmt"
    "io"
    "net/http"
)

const (
    APIKey = "your-api-key-here"
    BaseURL = "https://api.vidgo.ai"
)

type APIResponse struct {
    Code int      `json:"code"`
    Data TaskData `json:"data"`
}

type TaskData struct {
    TaskID       string      `json:"task_id"`
    Status       string      `json:"status"`
    Files        []MediaFile `json:"files"`
    CreatedTime  string      `json:"created_time"`
    Progress     int         `json:"progress"`
    ErrorMessage *string     `json:"error_message"`
}

type MediaFile struct {
    FileURL  string `json:"file_url"`
    FileType string `json:"file_type"`
}

func main() {
    taskID := "task-unified-1757165031-uyujaw3d"

    req, _ := http.NewRequest(
        "GET",
        fmt.Sprintf("%s/api/generate/status/%s", BaseURL, taskID),
        nil,
    )
    req.Header.Set("Authorization", "Bearer "+APIKey)

    client := &http.Client{}
    resp, err := client.Do(req)
    if err != nil {
        panic(err)
    }
    defer resp.Body.Close()

    body, _ := io.ReadAll(resp.Body)

    var apiResp APIResponse
    json.Unmarshal(body, &apiResp)

    data := apiResp.Data
    fmt.Printf("Status: %s\n", data.Status)
    fmt.Printf("Progress: %d%%\n", data.Progress)

    if data.Status == "finished" {
        for _, file := range data.Files {
            fmt.Printf("Generated file: %s\n", file.FileURL)
        }
    } else if data.Status == "failed" {
        fmt.Printf("Error: %s\n", *data.ErrorMessage)
    }
}
```

### Java

```java theme={null}
import java.net.http.*;
import java.net.URI;
import com.fasterxml.jackson.databind.ObjectMapper;
import com.fasterxml.jackson.databind.JsonNode;

public class VidgoAPI {
    private static final String API_KEY = "your-api-key-here";
    private static final String BASE_URL = "https://api.vidgo.ai";

    public static void main(String[] args) throws Exception {
        String taskId = "task-unified-1757165031-uyujaw3d";

        HttpClient client = HttpClient.newHttpClient();
        HttpRequest request = HttpRequest.newBuilder()
            .uri(URI.create(BASE_URL + "/api/generate/status/" + taskId))
            .header("Authorization", "Bearer " + API_KEY)
            .GET()
            .build();

        HttpResponse<String> response = client.send(
            request,
            HttpResponse.BodyHandlers.ofString()
        );

        ObjectMapper mapper = new ObjectMapper();
        JsonNode root = mapper.readTree(response.body());
        JsonNode data = root.get("data");

        String status = data.get("status").asText();
        int progress = data.get("progress").asInt();

        System.out.println("Status: " + status);
        System.out.println("Progress: " + progress + "%");

        if ("finished".equals(status)) {
            JsonNode files = data.get("files");
            files.forEach(file -> {
                System.out.println("Generated file: " + file.get("file_url").asText());
            });
        } else if ("failed".equals(status)) {
            System.out.println("Error: " + data.get("error_message").asText());
        }
    }
}
```

### PHP

```php theme={null}
<?php

$apiKey = 'your-api-key-here';
$taskId = 'task-unified-1757165031-uyujaw3d';
$baseUrl = 'https://api.vidgo.ai';

$ch = curl_init();
curl_setopt($ch, CURLOPT_URL, "$baseUrl/api/generate/status/$taskId");
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_HTTPHEADER, [
    "Authorization: Bearer $apiKey"
]);

$response = curl_exec($ch);
curl_close($ch);

$result = json_decode($response, true);
$data = $result['data'];

echo "Status: {$data['status']}\n";
echo "Progress: {$data['progress']}%\n";

if ($data['status'] === 'finished') {
    foreach ($data['files'] as $file) {
        echo "Generated file: {$file['file_url']}\n";
    }
} elseif ($data['status'] === 'failed') {
    echo "Error: {$data['error_message']}\n";
}
```

### Ruby

```ruby theme={null}
require 'net/http'
require 'json'
require 'uri'

api_key = 'your-api-key-here'
task_id = 'task-unified-1757165031-uyujaw3d'
base_url = 'https://api.vidgo.ai'

uri = URI("#{base_url}/api/generate/status/#{task_id}")
request = Net::HTTP::Get.new(uri)
request['Authorization'] = "Bearer #{api_key}"

response = Net::HTTP.start(uri.hostname, uri.port, use_ssl: true) do |http|
  http.request(request)
end

result = JSON.parse(response.body)
data = result['data']

puts "Status: #{data['status']}"
puts "Progress: #{data['progress']}%"

if data['status'] == 'finished'
  data['files'].each do |file|
    puts "Generated file: #{file['file_url']}"
  end
elsif data['status'] == 'failed'
  puts "Error: #{data['error_message']}"
end
```

## Polling Best Practices

When polling for task completion, follow these recommendations:

> [!TIP]
>
> **Poll Interval**: Check status every 2-5 seconds. Avoid polling more frequently to prevent rate limiting.

> [!TIP]
>
> **Timeout**: Implement timeouts to avoid infinite loops. Typical generation times:
>
> * Images: 15-60 seconds
> * Videos: 90-300 seconds

> [!TIP]
>
> **Exponential Backoff**: On rate limit (429) or server errors (5xx), implement exponential backoff (wait 1s, 2s, 4s, 8s...).

### Complete Polling Example

```python theme={null}
import requests
import time

API_KEY = "your-api-key-here"
BASE_URL = "https://api.vidgo.ai"

def wait_for_completion(task_id, timeout=600, poll_interval=2):
    """
    Wait for task completion with timeout

    Args:
        task_id: Task identifier
        timeout: Maximum wait time in seconds
        poll_interval: Seconds between checks

    Returns:
        Task data when complete

    Raises:
        TimeoutError: If timeout exceeded
        RuntimeError: If task failed
    """
    start_time = time.time()
    headers = {"Authorization": f"Bearer {API_KEY}"}

    while True:
        # Check timeout
        if time.time() - start_time > timeout:
            raise TimeoutError(f"Task did not complete within {timeout}s")

        # Get status
        response = requests.get(
            f"{BASE_URL}/api/generate/status/{task_id}",
            headers=headers
        )

        if response.status_code == 429:
            # Rate limited - wait longer
            time.sleep(poll_interval * 2)
            continue

        response.raise_for_status()
        result = response.json()
        data = result["data"]

        # Log progress
        elapsed = int(time.time() - start_time)
        print(f"[{elapsed}s] Status: {data['status']}, Progress: {data['progress']}%")

        # Check completion
        if data["status"] == "finished":
            print("Task completed successfully")
            return data
        elif data["status"] == "failed":
            raise RuntimeError(f"Task failed: {data['error_message']}")

        time.sleep(poll_interval)

# Usage
try:
    task_id = "task-unified-1757165031-uyujaw3d"
    result = wait_for_completion(task_id)

    for file in result["files"]:
        print(f"Download: {file['file_url']}")

except TimeoutError as e:
    print(f"Timeout: {e}")
except RuntimeError as e:
    print(f"Error: {e}")
```

## Error Responses

### HTTP Error Codes

| Code | Description | Action |
| - | - | - |
| 404 | Not Found | Task ID does not exist |
| 500 | Server Error | Retry with exponential backoff |

### Error Response Format

```json theme={null}
{
  "detail": "Task not found"
}
```

## Important Notes

> [!WARNING]
>
> **Content Validity**: File expiry depends on the storage provider. Download and save outputs that you need to retain.

> [!NOTE]
>
> **Credits**: Credits are deducted when the task is submitted. If the task fails, those credits are refunded automatically.

> [!TIP]
>
> **Webhooks**: Instead of polling, you can provide a `callback_url` when submitting tasks to receive automatic notifications. See [Webhook Callbacks](/api-manual/task-management/webhooks).

## Next Steps

> [card] [Submit Task](/api-manual/quickstart)
>
> Learn how to submit generation tasks

> [card] [Image Generation](/api-manual/image-series/gpt-4o-image)
>
> Generate images with GPT-4o

> [card] [Video Generation](/api-manual/video-series/overview)
>
> Create videos with Sora 2

> [card] [Quick Start](/api-manual/quickstart)
>
> Complete integration guide

## Additional generated files

Audio and 3D workflows use the same files array. file\_type may be audio, 3d, image, or other. The array retains model assets, previews, textures, animation files, and timestamp files. Each item may also include label, format, content\_type, file\_name, file\_size, and metadata. Process every returned file; metadata fields are optional.


# Webhook Callbacks
Source: https://docs.vidgo.ai/api-manual/task-management/webhooks

Receive automatic notifications when tasks complete

Instead of polling the status endpoint, you can provide a callback URL when submitting tasks to receive automatic notifications when generation completes.

## How It Works

1. **Submit with callback URL**: Include a `callback_url` in your task submission
2. **Receive notification**: When the task completes (success or failure), Vidgo API sends a POST request to your URL
3. **Process result**: Your server receives the complete task data including generated files

## Callback Request

When a task reaches `finished` or `failed`, Vidgo API sends a flat task object to your callback URL. Unlike the [status endpoint response](/api-manual/task-management/status), the callback body is not wrapped in `code` and `data`:

```json theme={null}
{
  "task_id": "task-unified-1757165031-uyujaw3d",
  "status": "finished",
  "files": [
    {
      "file_url": "https://file.goseaapp.com/images/image-abc123.jpg",
      "file_type": "image"
    }
  ],
  "created_time": "2025-11-12T10:30:00",
  "error_message": null
}
```

### Failed Task Callback

```json theme={null}
{
  "task_id": "task-unified-1757165031-uyujaw3d",
  "status": "failed",
  "files": [],
  "created_time": "2025-11-12T10:30:00",
  "error_message": "The prompt violates our content policy"
}
```

## Requirements

Your callback endpoint must meet the following requirements:

* **Public endpoint**: Use a publicly reachable HTTP(S) URL
* **Response timeout**: Respond within 30 seconds
* **Success response**: Should return HTTP 200-299 status code

## Retry Policy

If delivery times out, encounters a network error, or receives a `5xx` response:

* **Retry attempts**: Up to 5 automatic retries
* **Retry delays**: 60, 120, 240, 480, and 600 seconds between attempts
* **Final failure**: After all retries are exhausted, no further callback attempts are made
* **Client errors**: `4xx` responses are recorded and are not retried

> [!WARNING]
>
> If all retry attempts fail, you can still retrieve the results by polling the [status endpoint](/api-manual/task-management/status).

## Example Implementation

### Python (Flask)

```python theme={null}
from flask import Flask, request, jsonify

app = Flask(__name__)

@app.route('/webhook/generation-complete', methods=['POST'])
def handle_generation_complete():
    # Parse the callback payload
    task_data = request.get_json(silent=True) or {}

    task_id = task_data.get('task_id')
    status = task_data.get('status')

    if status == 'finished':
        # Task completed successfully
        files = task_data.get('files', [])
        for file in files:
            print(f"Generated file: {file['file_url']}")
            # Download and process the file
            # save_file(file['file_url'])

    elif status == 'failed':
        # Task failed
        error = task_data.get('error_message')
        print(f"Task {task_id} failed: {error}")
        # Handle the failure

    # Return 200 to acknowledge receipt
    return jsonify({"received": True}), 200

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=443, ssl_context='adhoc')
```

### Node.js (Express)

```javascript theme={null}
const express = require('express');
const app = express();

app.use(express.json());

app.post('/webhook/generation-complete', (req, res) => {
  const { task_id, status, files = [], error_message } = req.body;

  if (status === 'finished') {
    // Task completed successfully
    console.log(`Task ${task_id} completed`);
    files.forEach(file => {
      console.log(`Generated file: ${file.file_url}`);
      // Download and process the file
    });
  } else if (status === 'failed') {
    // Task failed
    console.log(`Task ${task_id} failed: ${error_message}`);
    // Handle the failure
  }

  // Return 200 to acknowledge receipt
  res.status(200).json({ received: true });
});

const https = require('https');
const fs = require('fs');

const options = {
  key: fs.readFileSync('private-key.pem'),
  cert: fs.readFileSync('certificate.pem')
};

https.createServer(options, app).listen(443, () => {
  console.log('Webhook server listening on port 443');
});
```

### Go

```go theme={null}
package main

import (
    "encoding/json"
    "fmt"
    "net/http"
)

type TaskData struct {
    TaskID       string      `json:"task_id"`
    Status       string      `json:"status"`
    Files        []MediaFile `json:"files"`
    CreatedTime  string      `json:"created_time"`
    ErrorMessage *string     `json:"error_message"`
}

type MediaFile struct {
    FileURL  string `json:"file_url"`
    FileType string `json:"file_type"`
}

func handleWebhook(w http.ResponseWriter, r *http.Request) {
    var taskData TaskData

    if err := json.NewDecoder(r.Body).Decode(&taskData); err != nil {
        http.Error(w, err.Error(), http.StatusBadRequest)
        return
    }

    if taskData.Status == "finished" {
        // Task completed successfully
        fmt.Printf("Task %s completed\n", taskData.TaskID)
        for _, file := range taskData.Files {
            fmt.Printf("Generated file: %s\n", file.FileURL)
            // Download and process the file
        }
    } else if taskData.Status == "failed" {
        // Task failed
        errorMessage := "unknown error"
        if taskData.ErrorMessage != nil {
            errorMessage = *taskData.ErrorMessage
        }
        fmt.Printf("Task %s failed: %s\n", taskData.TaskID, errorMessage)
        // Handle the failure
    }

    // Return 200 to acknowledge receipt
    w.WriteHeader(http.StatusOK)
    json.NewEncoder(w).Encode(map[string]bool{"received": true})
}

func main() {
    http.HandleFunc("/webhook/generation-complete", handleWebhook)

    // Use HTTPS
    err := http.ListenAndServeTLS(":443", "cert.pem", "key.pem", nil)
    if err != nil {
        panic(err)
    }
}
```

## Security Best Practices

> [!TIP]
>
> **Verify requests**: Consider adding a signature verification mechanism to ensure requests are from Vidgo API

> [!TIP]
>
> **Idempotency**: Design your webhook handler to be idempotent in case of duplicate deliveries

> [!TIP]
>
> **Async processing**: Process the callback asynchronously and return 200 quickly to avoid timeouts

> [!TIP]
>
> **Logging**: Log all webhook requests for debugging and monitoring

## Testing

### Using ngrok for Local Testing

During development, you can use [ngrok](https://ngrok.com) to expose your local server:

```bash theme={null}
# Start your local webhook server on port 3000
node webhook-server.js

# In another terminal, start ngrok
ngrok http 3000

# Use the HTTPS URL provided by ngrok as your callback_url
# Example: https://abc123.ngrok.io/webhook/generation-complete
```

## Webhook vs Polling

| Method | Best For | Pros | Cons |
| - | - | - | - |
| **Webhooks** | Production systems | Real-time notifications, no polling overhead | Requires public endpoint, harder to debug |
| **Polling** | Development, simple integrations | Easy to implement, no server required | Higher latency, consumes more resources |

> [!NOTE]
>
> For production systems handling high volumes, webhooks are recommended to reduce API calls and get instant notifications.

## Next Steps

> [card] [Task Status](/api-manual/task-management/status)
>
> Learn about polling task status

> [card] [Image Generation](/api-manual/image-series/gpt-4o-image)
>
> Generate images with GPT-4o

> [card] [Video Generation](/api-manual/video-series/overview)
>
> Create videos with Sora 2

> [card] [Quick Start](/api-manual/quickstart)
>
> Back to Quick Start guide

## Additional generated files

Audio and 3D workflows use the same files array. file\_type may be audio, 3d, image, or other. The array retains model assets, previews, textures, animation files, and timestamp files. Each item may also include label, format, content\_type, file\_name, file\_size, and metadata. Process every returned file; metadata fields are optional.


# Gemini Omni 1.1 Flash Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-gemini-omni-1-1-flash-image-to-video

api-manual/video-series/google-gemini-omni-1-1-flash-image-to-video.json POST /api/generate/submit
Animate a still image with Gemini Omni 1.1 Flash, using text direction to develop a short video scene at up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate a still image with Gemini Omni 1.1 Flash, using text direction to develop a short video scene at up to 4K.

## Models

* **google/gemini-omni-1.1-flash/image-to-video** - Generate video from image input.

## Features

* Generate video from one or two input images.
* Choose from `360p`, `720p`, `1080p`, `4k` output resolutions.
* Supports landscape (`16:9`) and portrait (`9:16`) framing.

## Duration Options

* **4, 6, 8, 10 seconds** - Set `duration` to one of these values. Default: **8 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Image URLs

* `image_urls`: Required. Provide 1-2 image URLs.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `resolution`: `360p`, `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`. Default: `16:9`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Gemini Omni 1.1 Flash Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-gemini-omni-1-1-flash-reference-to-video

api-manual/video-series/google-gemini-omni-1-1-flash-reference-to-video.json POST /api/generate/submit
Use visual references to guide new scenes with Gemini Omni 1.1 Flash, combining image input and text direction in video output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Use visual references to guide new scenes with Gemini Omni 1.1 Flash, combining image input and text direction in video output up to 4K.

## Models

* **google/gemini-omni-1.1-flash/reference-to-video** - Generate a new video guided by reference material and a prompt.

## Features

* Guide video generation with one to seven reference images.
* Choose from `360p`, `720p`, `1080p`, `4k` output resolutions.
* Supports landscape (`16:9`) and portrait (`9:16`) framing.

## Duration Options

* **4, 6, 8, 10 seconds** - Set `duration` to one of these values. Default: **8 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Reference Inputs

* `reference_image_urls`: Required; provide 1-7 images.

### Resolution

* `resolution`: `360p`, `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`. Default: `16:9`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Gemini Omni 1.1 Flash Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-gemini-omni-1-1-flash-text-to-video

api-manual/video-series/google-gemini-omni-1-1-flash-text-to-video.json POST /api/generate/submit
Turn written scene concepts into short videos with Gemini Omni 1.1 Flash, with landscape or portrait framing and output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn written scene concepts into short videos with Gemini Omni 1.1 Flash, with landscape or portrait framing and output up to 4K.

## Models

* **google/gemini-omni-1.1-flash/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `360p`, `720p`, `1080p`, `4k` output resolutions.
* Supports landscape (`16:9`) and portrait (`9:16`) framing.

## Duration Options

* **4, 6, 8, 10 seconds** - Set `duration` to one of these values. Default: **8 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Resolution

* `resolution`: `360p`, `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`. Default: `16:9`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Gemini Omni 1.1 Flash Video Edit
Source: https://docs.vidgo.ai/api-manual/video-series/google-gemini-omni-1-1-flash-video-edit

api-manual/video-series/google-gemini-omni-1-1-flash-video-edit.json POST /api/generate/submit
Rework existing video with Gemini Omni 1.1 Flash, using written instructions to guide changes to the scene with output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Rework existing video with Gemini Omni 1.1 Flash, using written instructions to guide changes to the scene with output up to 4K.

## Models

* **google/gemini-omni-1.1-flash/video-edit** - Edit an existing video using written instructions.

## Features

* Edit an existing video using written instructions.
* Optionally add up to five reference images to guide the edit.
* Choose from `360p`, `720p`, `1080p`, `4k` output resolutions.
* Supports landscape (`16:9`) and portrait (`9:16`) framing.

## Duration Options

* **Source video: 3-10 seconds** - Output duration follows the reference video.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes to the video.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Reference Inputs

* `reference_image_urls`: Optional; provide up to 5 images.
* Omit `reference_image_urls` or send an empty array when no reference images are needed.

### Source Video

* `reference_video_urls`: Required. Provide an array containing exactly one publicly accessible source video URL.
* The source video must be directly downloadable over HTTP(S) and no larger than **100 MiB** for duration inspection.

### Resolution

* `resolution`: `360p`, `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`. Default: `16:9`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Omni Flash Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-gemini-omni-flash-image-to-video

api-manual/video-series/google-gemini-omni-flash-image-to-video.json POST /api/generate/submit
Animate a still image with Gemini Omni Flash, using text direction to develop a short video scene at up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate a still image with Gemini Omni Flash, using text direction to develop a short video scene at up to 4K.

## Models

* **google/gemini-omni-flash/image-to-video** - Generate video from image input.

## Features

* Animate a single reference image.
* Choose from `720p`, `1080p`, `4k` output resolutions.
* Supports landscape (`16:9`) and portrait (`9:16`) framing.

## Duration Options

* **4, 6, 8, 10 seconds** - Set `duration` to one of these values. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `resolution`: `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`. Default: `16:9`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Omni Flash Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-gemini-omni-flash-reference-to-video

api-manual/video-series/google-gemini-omni-flash-reference-to-video.json POST /api/generate/submit
Use visual references to guide new scenes with Gemini Omni Flash, combining image input and text direction in video output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Use visual references to guide new scenes with Gemini Omni Flash, combining image input and text direction in video output up to 4K.

## Models

* **google/gemini-omni-flash/reference-to-video** - Generate a new video guided by reference material and a prompt.

## Features

* Combine exactly three reference images with a prompt to guide the video.
* Choose from `720p`, `1080p`, `4k` output resolutions.
* Supports landscape (`16:9`) and portrait (`9:16`) framing.

## Duration Options

* **4, 6, 8, 10 seconds** - Set `duration` to one of these values. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Image URLs

* `image_urls`: Required. Provide exactly 3 image URLs.
* Use `image_urls` for this workflow.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `resolution`: `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`. Default: `16:9`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Omni Flash Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-gemini-omni-flash-text-to-video

api-manual/video-series/google-gemini-omni-flash-text-to-video.json POST /api/generate/submit
Turn written scene concepts into short videos with Gemini Omni Flash, with landscape or portrait framing and output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn written scene concepts into short videos with Gemini Omni Flash, with landscape or portrait framing and output up to 4K.

## Models

* **google/gemini-omni-flash/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p`, `4k` output resolutions.
* Supports landscape (`16:9`) and portrait (`9:16`) framing.

## Duration Options

* **4, 6, 8, 10 seconds** - Set `duration` to one of these values. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Resolution

* `resolution`: `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`. Default: `16:9`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Omni Flash Video Edit
Source: https://docs.vidgo.ai/api-manual/video-series/google-gemini-omni-flash-video-edit

api-manual/video-series/google-gemini-omni-flash-video-edit.json POST /api/generate/submit
Rework existing video with Gemini Omni Flash, using written instructions to guide changes to the scene with output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Rework existing video with Gemini Omni Flash, using written instructions to guide changes to the scene with output up to 4K.

## Models

* **google/gemini-omni-flash/video-edit** - Edit an existing video using written instructions.

## Features

* Edit an existing video using written instructions.
* Choose from `720p`, `1080p`, `4k` output resolutions.
* Supports landscape (`16:9`) and portrait (`9:16`) framing.

## Duration Options

* Provide one source video in `video_urls`. This workflow does not impose a source-video duration limit.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes to the video.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Source Video

* `video_urls`: Required. Provide an array containing exactly one publicly accessible source video URL.

### Resolution

* `resolution`: `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`. Default: `16:9`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Video Models
Source: https://docs.vidgo.ai/api-manual/video-series/overview

All video generation models available on Vidgo API, with their model IDs.

Every video model is called through `POST /api/generate/submit`. Send a model ID in `model`, the model's parameters in `input`, and an optional `callback_url`. The request returns a `task_id` right away and the task runs asynchronously.

## Workflow

1. Submit a task to `POST /api/generate/submit` with a model ID from the table below.
2. Poll [Get Task Status](/api-manual/task-management/status) with the returned `task_id`, or receive the result at `callback_url` (see [Webhook Callbacks](/api-manual/task-management/webhooks)).
3. Download the files in `files` when the status is `finished`.

Some models expose one model ID per mode, such as text-to-video, image-to-video, or reference-to-video. Others use a single ID and choose the mode from the inputs you send; see each model page.

## Models

| Model | Provider | Model IDs |
| - | - | - |
| [Veo 3.1 Official](/api-manual/video-series/veo-3-1-official) | Google | `veo3.1-fast-official` `veo3.1-lite-official` `veo3.1-quality-official` |
| [Veo 3.1](/api-manual/video-series/veo-3-1) | Google | `veo3.1-fast` `veo3.1-quality` |
| [Veo 3.1 Fast Text to Video](/api-manual/video-series/veo3-1-fast-text-to-video) | Google | `veo3.1-fast/text-to-video` |
| [Veo 3.1 Fast Image to Video](/api-manual/video-series/veo3-1-fast-image-to-video) | Google | `veo3.1-fast/image-to-video` |
| [Veo 3.1 Fast Reference to Video](/api-manual/video-series/veo3-1-fast-reference-to-video) | Google | `veo3.1-fast/reference-to-video` |
| [Veo 3.1 Lite Text to Video](/api-manual/video-series/veo3-1-lite-text-to-video) | Google | `veo3.1-lite/text-to-video` |
| [Veo 3.1 Quality Text to Video](/api-manual/video-series/veo3-1-text-to-video) | Google | `veo3.1/text-to-video` |
| [Veo 3.1 Quality Image to Video](/api-manual/video-series/veo3-1-image-to-video) | Google | `veo3.1/image-to-video` |
| [Gemini Omni 1.1 Flash Text to Video](/api-manual/video-series/google-gemini-omni-1-1-flash-text-to-video) | Google | `google/gemini-omni-1.1-flash/text-to-video` |
| [Gemini Omni 1.1 Flash Image to Video](/api-manual/video-series/google-gemini-omni-1-1-flash-image-to-video) | Google | `google/gemini-omni-1.1-flash/image-to-video` |
| [Gemini Omni 1.1 Flash Reference to Video](/api-manual/video-series/google-gemini-omni-1-1-flash-reference-to-video) | Google | `google/gemini-omni-1.1-flash/reference-to-video` |
| [Gemini Omni 1.1 Flash Video Edit](/api-manual/video-series/google-gemini-omni-1-1-flash-video-edit) | Google | `google/gemini-omni-1.1-flash/video-edit` |
| [Omni Flash Text to Video](/api-manual/video-series/google-gemini-omni-flash-text-to-video) | Google | `google/gemini-omni-flash/text-to-video` |
| [Omni Flash Image to Video](/api-manual/video-series/google-gemini-omni-flash-image-to-video) | Google | `google/gemini-omni-flash/image-to-video` |
| [Omni Flash Reference to Video](/api-manual/video-series/google-gemini-omni-flash-reference-to-video) | Google | `google/gemini-omni-flash/reference-to-video` |
| [Omni Flash Video Edit](/api-manual/video-series/google-gemini-omni-flash-video-edit) | Google | `google/gemini-omni-flash/video-edit` |
| [Veo 3.1 Official Text to Video](/api-manual/video-series/google-veo3-1-text-to-video) | Google | `google/veo3.1/text-to-video` |
| [Veo 3.1 Official Image to Video](/api-manual/video-series/google-veo3-1-image-to-video) | Google | `google/veo3.1/image-to-video` |
| [Veo 3.1 Fast Text to Video](/api-manual/video-series/google-veo3-1-fast-text-to-video) | Google | `google/veo3.1-fast/text-to-video` |
| [Veo 3.1 Fast Image to Video](/api-manual/video-series/google-veo3-1-fast-image-to-video) | Google | `google/veo3.1-fast/image-to-video` |
| [Veo 3.1 Lite Text to Video](/api-manual/video-series/google-veo3-1-lite-text-to-video) | Google | `google/veo3.1-lite/text-to-video` |
| [Veo 3.1 Lite Image to Video](/api-manual/video-series/google-veo3-1-lite-image-to-video) | Google | `google/veo3.1-lite/image-to-video` |
| [FLUX 3 Text to Video](/api-manual/video-series/blackforestlabs-flux-3-text-to-video) | Flux | `blackforestlabs/flux-3/text-to-video` |
| [FLUX 3 Image to Video](/api-manual/video-series/blackforestlabs-flux-3-image-to-video) | Flux | `blackforestlabs/flux-3/image-to-video` |
| [FLUX 3 First Last Frame to Video](/api-manual/video-series/blackforestlabs-flux-3-first-last-frame-to-video) | Flux | `blackforestlabs/flux-3/first-last-frame-to-video` |
| [FLUX 3 Extend Video](/api-manual/video-series/blackforestlabs-flux-3-extend-video) | Flux | `blackforestlabs/flux-3/extend-video` |
| [FLUX 3 Keyframes to Video](/api-manual/video-series/blackforestlabs-flux-3-keyframes-to-video) | Flux | `blackforestlabs/flux-3/keyframes-to-video` |
| [FLUX 3 Max Text to Video](/api-manual/video-series/blackforestlabs-flux-3-max-text-to-video) | Flux | `blackforestlabs/flux-3-max/text-to-video` |
| [FLUX 3 Max Image to Video](/api-manual/video-series/blackforestlabs-flux-3-max-image-to-video) | Flux | `blackforestlabs/flux-3-max/image-to-video` |
| [FLUX 3 Max Keyframes to Video](/api-manual/video-series/blackforestlabs-flux-3-max-keyframes-to-video) | Flux | `blackforestlabs/flux-3-max/keyframes-to-video` |
| [Kling 3.0](/api-manual/video-series/kling-3-0) | Kling | `kling-3.0/pro` `kling-3.0/standard` |
| [Kling 3.0 Standard Text to Video](/api-manual/video-series/kwaivgi-kling-v3-0-std-text-to-video) | Kling | `kwaivgi/kling-v3.0-std/text-to-video` |
| [Kling 3.0 Standard Image to Video](/api-manual/video-series/kwaivgi-kling-v3-0-std-image-to-video) | Kling | `kwaivgi/kling-v3.0-std/image-to-video` |
| [Kling 3.0 Pro Text to Video](/api-manual/video-series/kwaivgi-kling-v3-0-pro-text-to-video) | Kling | `kwaivgi/kling-v3.0-pro/text-to-video` |
| [Kling 3.0 Pro Image to Video](/api-manual/video-series/kwaivgi-kling-v3-0-pro-image-to-video) | Kling | `kwaivgi/kling-v3.0-pro/image-to-video` |
| [Kling 3.0 4K Text to Video](/api-manual/video-series/kwaivgi-kling-v3-0-4k-text-to-video) | Kling | `kwaivgi/kling-v3.0-4k/text-to-video` |
| [Kling 3.0 4K Image to Video](/api-manual/video-series/kwaivgi-kling-v3-0-4k-image-to-video) | Kling | `kwaivgi/kling-v3.0-4k/image-to-video` |
| [Kling 3.0 Motion Control](/api-manual/video-series/kling-3-0-motion-control) | Kling | `kling-3.0-motion-control` |
| [Kling 2.6 Pro Text to Video](/api-manual/video-series/kwaivgi-kling-v2-6-pro-text-to-video) | Kling | `kwaivgi/kling-v2.6-pro/text-to-video` |
| [Kling 2.6 Pro Image to Video](/api-manual/video-series/kwaivgi-kling-v2-6-pro-image-to-video) | Kling | `kwaivgi/kling-v2.6-pro/image-to-video` |
| [Kling 2.6 Motion Control](/api-manual/video-series/kling-2.6-motion-control) | Kling | `kling-2.6-motion-control` |
| [Kling 2.5 Turbo Pro Text to Video](/api-manual/video-series/kwaivgi-kling-v2-5-turbo-pro-text-to-video) | Kling | `kwaivgi/kling-v2.5-turbo-pro/text-to-video` |
| [Kling 2.5 Turbo Pro Image to Video](/api-manual/video-series/kwaivgi-kling-v2-5-turbo-pro-image-to-video) | Kling | `kwaivgi/kling-v2.5-turbo-pro/image-to-video` |
| [Kling 2.1](/api-manual/video-series/kling-2-1) | Kling | `kling-2.1/pro` `kling-2.1/standard` |
| [Kling 2.1 Image to Video](/api-manual/video-series/kwaivgi-kling-video-v2-1-standard-image-to-video) | Kling | `kwaivgi/kling-video/v2.1/standard/image-to-video` |
| [Kling 2.1 Pro Image to Video](/api-manual/video-series/kwaivgi-kling-video-v2-1-pro-image-to-video) | Kling | `kwaivgi/kling-video/v2.1/pro/image-to-video` |
| [Kling 1.6](/api-manual/video-series/kling-1-6) | Kling | `kling-1.6/pro` `kling-1.6/standard` |
| [Kling 1.6 Standard Text to Video](/api-manual/video-series/kwaivgi-kling-video-v1-6-standard-text-to-video) | Kling | `kwaivgi/kling-video/v1.6/standard/text-to-video` |
| [Kling 1.6 Standard Image to Video](/api-manual/video-series/kwaivgi-kling-video-v1-6-standard-image-to-video) | Kling | `kwaivgi/kling-video/v1.6/standard/image-to-video` |
| [Kling 1.6 Pro Text to Video](/api-manual/video-series/kwaivgi-kling-video-v1-6-pro-text-to-video) | Kling | `kwaivgi/kling-video/v1.6/pro/text-to-video` |
| [Kling 1.6 Pro Image to Video](/api-manual/video-series/kwaivgi-kling-video-v1-6-pro-image-to-video) | Kling | `kwaivgi/kling-video/v1.6/pro/image-to-video` |
| [Kling 3.0 Turbo Standard Text to Video](/api-manual/video-series/kwaivgi-kling-v3-turbo-std-text-to-video) | Kling | `kwaivgi/kling-v3-turbo-std/text-to-video` |
| [Kling 3.0 Turbo Standard Image to Video](/api-manual/video-series/kwaivgi-kling-v3-turbo-std-image-to-video) | Kling | `kwaivgi/kling-v3-turbo-std/image-to-video` |
| [Kling 3.0 Turbo Pro Text to Video](/api-manual/video-series/kwaivgi-kling-v3-turbo-pro-text-to-video) | Kling | `kwaivgi/kling-v3-turbo-pro/text-to-video` |
| [Kling 3.0 Turbo Pro Image to Video](/api-manual/video-series/kwaivgi-kling-v3-turbo-pro-image-to-video) | Kling | `kwaivgi/kling-v3-turbo-pro/image-to-video` |
| [Kling 3.0 Standard Motion Control](/api-manual/video-series/kwaivgi-kling-v3-0-std-motion-control) | Kling | `kwaivgi/kling-v3.0-std/motion-control` |
| [Kling 3.0 Pro Motion Control](/api-manual/video-series/kwaivgi-kling-v3-0-pro-motion-control) | Kling | `kwaivgi/kling-v3.0-pro/motion-control` |
| [Kling 2.6 Standard Motion Control](/api-manual/video-series/kwaivgi-kling-v2-6-std-motion-control) | Kling | `kwaivgi/kling-v2.6-std/motion-control` |
| [Kling 2.6 Pro Motion Control](/api-manual/video-series/kwaivgi-kling-v2-6-pro-motion-control) | Kling | `kwaivgi/kling-v2.6-pro/motion-control` |
| [Kling Avatar 2.0 Standard](/api-manual/video-series/kwaivgi-kling-video-ai-avatar-v2-standard) | Kling | `kwaivgi/kling-video/ai-avatar/v2/standard` |
| [Kling Avatar 2.0 Pro](/api-manual/video-series/kwaivgi-kling-video-ai-avatar-v2-pro) | Kling | `kwaivgi/kling-video/ai-avatar/v2/pro` |
| [Kling O3 Standard Text To Video](/api-manual/video-series/kwaivgi-kling-video-o3-std-text-to-video) | Kling | `kwaivgi/kling-video-o3-std/text-to-video` |
| [Kling O3 Standard Image To Video](/api-manual/video-series/kwaivgi-kling-video-o3-std-image-to-video) | Kling | `kwaivgi/kling-video-o3-std/image-to-video` |
| [Kling O3 Pro Text To Video](/api-manual/video-series/kwaivgi-kling-video-o3-pro-text-to-video) | Kling | `kwaivgi/kling-video-o3-pro/text-to-video` |
| [Kling O3 Pro Image To Video](/api-manual/video-series/kwaivgi-kling-video-o3-pro-image-to-video) | Kling | `kwaivgi/kling-video-o3-pro/image-to-video` |
| [Kling O3 4K Text To Video](/api-manual/video-series/kwaivgi-kling-video-o3-4k-text-to-video) | Kling | `kwaivgi/kling-video-o3-4k/text-to-video` |
| [Kling O3 4K Image To Video](/api-manual/video-series/kwaivgi-kling-video-o3-4k-image-to-video) | Kling | `kwaivgi/kling-video-o3-4k/image-to-video` |
| [Kling 4.0 Text to Video](/api-manual/video-series/kwaivgi-kling-v4-0-text-to-video) | Kling | `kwaivgi/kling-v4.0/text-to-video` |
| [Kling 4.0 Image to Video](/api-manual/video-series/kwaivgi-kling-v4-0-image-to-video) | Kling | `kwaivgi/kling-v4.0/image-to-video` |
| [Kling 4.0 Reference to Video](/api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video) | Kling | `kwaivgi/kling-v4.0/reference-to-video` |
| [Kling 4.0 Video Edit](/api-manual/video-series/kwaivgi-kling-v4-0-video-edit) | Kling | `kwaivgi/kling-v4.0/video-edit` |
| [Grok Imagine Text to Video](/api-manual/video-series/xai-grok-imagine-video-text-to-video) | Grok (xAI) | `xai/grok-imagine-video/text-to-video` |
| [Grok Imagine Image to Video](/api-manual/video-series/xai-grok-imagine-video-image-to-video) | Grok (xAI) | `xai/grok-imagine-video/image-to-video` |
| [Grok Imagine Video 1.5 Image to Video](/api-manual/video-series/xai-grok-imagine-video-v1-5-image-to-video) | Grok (xAI) | `xai/grok-imagine-video-v1.5/image-to-video` |
| [Happy Horse](/api-manual/video-series/happy-horse) | Alibaba | `happy-horse` |
| [Happy Horse Text to Video](/api-manual/video-series/alibaba-happyhorse-1-0-text-to-video) | Alibaba | `alibaba/happyhorse-1.0/text-to-video` |
| [Happy Horse Image to Video](/api-manual/video-series/alibaba-happyhorse-1-0-image-to-video) | Alibaba | `alibaba/happyhorse-1.0/image-to-video` |
| [Happy Horse Reference to Video](/api-manual/video-series/alibaba-happyhorse-1-0-reference-to-video) | Alibaba | `alibaba/happyhorse-1.0/reference-to-video` |
| [Happy Horse Video Edit](/api-manual/video-series/alibaba-happyhorse-1-0-video-edit) | Alibaba | `alibaba/happyhorse-1.0/video-edit` |
| [Happy Horse 1.1 Text to Video](/api-manual/video-series/alibaba-happyhorse-1-1-text-to-video) | Alibaba | `alibaba/happyhorse-1.1/text-to-video` |
| [Happy Horse 1.1 Image to Video](/api-manual/video-series/alibaba-happyhorse-1-1-image-to-video) | Alibaba | `alibaba/happyhorse-1.1/image-to-video` |
| [Happy Horse 1.1 Reference to Video](/api-manual/video-series/alibaba-happyhorse-1-1-reference-to-video) | Alibaba | `alibaba/happyhorse-1.1/reference-to-video` |
| [Wan 2.6](/api-manual/video-series/wan-2-6) | Alibaba | `wan2.6-image-to-video` `wan2.6-text-to-video` `wan2.6-video-to-video` |
| [Wan 2.5 Text-to-Video](/api-manual/video-series/wan2.5-text-to-video) | Alibaba | `wan2.5-text-to-video` |
| [Wan 2.5 Image-to-Video](/api-manual/video-series/wan2.5-image-to-video) | Alibaba | `wan2.5-image-to-video` |
| [Wan 2.5 Text to Video](/api-manual/video-series/alibaba-wan-2-5-text-to-video) | Alibaba | `alibaba/wan-2.5/text-to-video` |
| [Wan 2.5 Image to Video](/api-manual/video-series/alibaba-wan-2-5-image-to-video) | Alibaba | `alibaba/wan-2.5/image-to-video` |
| [Wan 2.2 Text-to-Video Fast](/api-manual/video-series/wan2.2-text-to-video-fast) | Alibaba | `wan2.2-text-to-video-fast` |
| [Wan 2.2 Image-to-Video Fast](/api-manual/video-series/wan2.2-image-to-video-fast) | Alibaba | `wan2.2-image-to-video-fast` |
| [Wan Animate Move](/api-manual/video-series/alibaba-wan-v2-2-14b-animate-move) | Alibaba | `alibaba/wan/v2.2-14b/animate/move` |
| [Wan Animate Replace](/api-manual/video-series/alibaba-wan-v2-2-14b-animate-replace) | Alibaba | `alibaba/wan/v2.2-14b/animate/replace` |
| [Wan 3.0 Text to Video](/api-manual/video-series/alibaba-wan-3-0-text-to-video) | Alibaba | `alibaba/wan-3.0/text-to-video` |
| [Wan 3.0 Image to Video](/api-manual/video-series/alibaba-wan-3-0-image-to-video) | Alibaba | `alibaba/wan-3.0/image-to-video` |
| [Wan 3.0 Reference to Video](/api-manual/video-series/alibaba-wan-3-0-reference-to-video) | Alibaba | `alibaba/wan-3.0/reference-to-video` |
| [Wan 3.0 Video Prime Text to Video](/api-manual/video-series/alibaba-wan-3-0-prime-text-to-video) | Alibaba | `alibaba/wan-3.0-prime/text-to-video` |
| [Wan 3.0 Video Prime Image to Video](/api-manual/video-series/alibaba-wan-3-0-prime-image-to-video) | Alibaba | `alibaba/wan-3.0-prime/image-to-video` |
| [Wan 3.0 Video Prime Reference to Video](/api-manual/video-series/alibaba-wan-3-0-prime-reference-to-video) | Alibaba | `alibaba/wan-3.0-prime/reference-to-video` |
| [Wan 2.7 Video Text to Video](/api-manual/video-series/alibaba-wan-2-7-text-to-video) | Alibaba | `alibaba/wan-2.7/text-to-video` |
| [Wan 2.7 Video Image to Video](/api-manual/video-series/alibaba-wan-2-7-image-to-video) | Alibaba | `alibaba/wan-2.7/image-to-video` |
| [Wan 2.7 Video Reference to Video](/api-manual/video-series/alibaba-wan-2-7-reference-to-video) | Alibaba | `alibaba/wan-2.7/reference-to-video` |
| [Wan 2.7 Video Edit](/api-manual/video-series/alibaba-wan-2-7-video-edit) | Alibaba | `alibaba/wan-2.7/video-edit` |
| [Wan 2.6 Text to Video](/api-manual/video-series/alibaba-wan-2-6-text-to-video) | Alibaba | `alibaba/wan-2.6/text-to-video` |
| [Wan 2.6 Image to Video](/api-manual/video-series/alibaba-wan-2-6-image-to-video) | Alibaba | `alibaba/wan-2.6/image-to-video` |
| [Wan 2.6 Reference to Video](/api-manual/video-series/alibaba-wan-2-6-reference-to-video) | Alibaba | `alibaba/wan-2.6/reference-to-video` |
| [Hailuo 2.3](/api-manual/video-series/hailuo-2-3) | MiniMax / Hailuo | `hailuo-2.3` |
| [Hailuo 2.3 Text to Video](/api-manual/video-series/minimax-hailuo-2-3-standard-text-to-video) | MiniMax / Hailuo | `minimax/hailuo-2.3/standard/text-to-video` |
| [Hailuo 2.3 Image to Video](/api-manual/video-series/minimax-hailuo-2-3-standard-image-to-video) | MiniMax / Hailuo | `minimax/hailuo-2.3/standard/image-to-video` |
| [Hailuo 2.3 Pro Text to Video](/api-manual/video-series/minimax-hailuo-2-3-pro-text-to-video) | MiniMax / Hailuo | `minimax/hailuo-2.3/pro/text-to-video` |
| [Hailuo 2.3 Pro Image to Video](/api-manual/video-series/minimax-hailuo-2-3-pro-image-to-video) | MiniMax / Hailuo | `minimax/hailuo-2.3/pro/image-to-video` |
| [Hailuo 02](/api-manual/video-series/hailuo-02) | MiniMax / Hailuo | `hailuo-02` `hailuo-02-pro` |
| [MiniMax H3 Text to Video](/api-manual/video-series/minimax-h3-text-to-video) | MiniMax / Hailuo | `minimax/h3/text-to-video` |
| [MiniMax H3 Image to Video](/api-manual/video-series/minimax-h3-image-to-video) | MiniMax / Hailuo | `minimax/h3/image-to-video` |
| [MiniMax H3 Reference to Video](/api-manual/video-series/minimax-h3-reference-to-video) | MiniMax / Hailuo | `minimax/h3/reference-to-video` |
| [MiniMax H3 Max Text to Video](/api-manual/video-series/minimax-h3-max-text-to-video) | MiniMax / Hailuo | `minimax/h3-max/text-to-video` |
| [MiniMax H3 Max Image to Video](/api-manual/video-series/minimax-h3-max-image-to-video) | MiniMax / Hailuo | `minimax/h3-max/image-to-video` |
| [MiniMax H3 Max Reference to Video](/api-manual/video-series/minimax-h3-max-reference-to-video) | MiniMax / Hailuo | `minimax/h3-max/reference-to-video` |
| [MiniMax H3 Max Turbo Text to Video](/api-manual/video-series/minimax-h3-max-turbo-text-to-video) | MiniMax / Hailuo | `minimax/h3-max-turbo/text-to-video` |
| [MiniMax H3 Max Turbo Image to Video](/api-manual/video-series/minimax-h3-max-turbo-image-to-video) | MiniMax / Hailuo | `minimax/h3-max-turbo/image-to-video` |
| [Hailuo 02 Text to Video](/api-manual/video-series/minimax-hailuo-02-standard-text-to-video) | MiniMax / Hailuo | `minimax/hailuo-02/standard/text-to-video` |
| [Hailuo 02 Image to Video](/api-manual/video-series/minimax-hailuo-02-standard-image-to-video) | MiniMax / Hailuo | `minimax/hailuo-02/standard/image-to-video` |
| [Hailuo 02 Pro Text to Video](/api-manual/video-series/minimax-hailuo-02-pro-text-to-video) | MiniMax / Hailuo | `minimax/hailuo-02/pro/text-to-video` |
| [Hailuo 02 Pro Image to Video](/api-manual/video-series/minimax-hailuo-02-pro-image-to-video) | MiniMax / Hailuo | `minimax/hailuo-02/pro/image-to-video` |
| [Runway Gen-4.5](/api-manual/video-series/runway-gen-4-5) | Runway | `runway-gen-4.5` |
| [Seedance 2](/api-manual/video-series/seedance-2) | Seedance | `seedance-2` `seedance-2-fast` |
| [Seedance 2.5 Text to Video](/api-manual/video-series/seedance-2-5-text-to-video) | Seedance | `seedance-2.5/text-to-video` |
| [Seedance 2.5 Image to Video](/api-manual/video-series/seedance-2-5-image-to-video) | Seedance | `seedance-2.5/image-to-video` |
| [Seedance 2.5 Reference to Video](/api-manual/video-series/seedance-2-5-reference-to-video) | Seedance | `seedance-2.5/reference-to-video` |
| [Seedance 2.0 Text to Video](/api-manual/video-series/seedance-2-0-text-to-video) | Seedance | `seedance-2` `seedance-2.0/text-to-video` |
| [Seedance 2.0 Image to Video](/api-manual/video-series/seedance-2-0-image-to-video) | Seedance | `seedance-2` `seedance-2.0/image-to-video` |
| [Seedance 2.0 Reference to Video](/api-manual/video-series/seedance-2-0-reference-to-video) | Seedance | `seedance-2` `seedance-2.0/reference-to-video` |
| [Seedance 2.0 Fast Text to Video](/api-manual/video-series/seedance-2-0-fast-text-to-video) | Seedance | `seedance-2-fast` `seedance-2.0-fast/text-to-video` |
| [Seedance 2.0 Fast Image to Video](/api-manual/video-series/seedance-2-0-fast-image-to-video) | Seedance | `seedance-2-fast` `seedance-2.0-fast/image-to-video` |
| [Seedance 2.0 Fast Reference to Video](/api-manual/video-series/seedance-2-0-fast-reference-to-video) | Seedance | `seedance-2-fast` `seedance-2.0-fast/reference-to-video` |
| [Seedance 2.0 Mini Text to Video](/api-manual/video-series/seedance-2-0-mini-text-to-video) | Seedance | `seedance-2.0-mini/text-to-video` |
| [Seedance 2.0 Mini Image to Video](/api-manual/video-series/seedance-2-0-mini-image-to-video) | Seedance | `seedance-2.0-mini/image-to-video` |
| [Seedance 2.0 Mini Reference to Video](/api-manual/video-series/seedance-2-0-mini-reference-to-video) | Seedance | `seedance-2.0-mini/reference-to-video` |
| [Seedance 1.5 Pro](/api-manual/video-series/seedance-1-5-pro) | Seedance | `seedance-1.5-pro` |
| [Seedance 1.5 Pro Text to Video](/api-manual/video-series/bytedance-seedance-v1-5-pro-text-to-video) | Seedance | `bytedance/seedance-v1.5-pro/text-to-video` |
| [Seedance 1.5 Pro Image to Video](/api-manual/video-series/bytedance-seedance-v1-5-pro-image-to-video) | Seedance | `bytedance/seedance-v1.5-pro/image-to-video` |
| [Seedance 1.0 Pro](/api-manual/video-series/seedance-1.0-pro) | Seedance | `seedance-1.0-pro` |
| [Seedance 1.0 Pro Text to Video](/api-manual/video-series/bytedance-seedance-v1-pro-text-to-video) | Seedance | `bytedance/seedance/v1/pro/text-to-video` |
| [Seedance 1.0 Pro Image to Video](/api-manual/video-series/bytedance-seedance-v1-pro-image-to-video) | Seedance | `bytedance/seedance/v1/pro/image-to-video` |
| [Vidu Q4 Image to Video](/api-manual/video-series/vidu-q4-image-to-video) | Vidu | `vidu/q4/image-to-video` |
| [Vidu Q4 Reference to Video](/api-manual/video-series/vidu-q4-reference-to-video) | Vidu | `vidu/q4/reference-to-video` |
| [OmniHuman 2.0 Audio to Video](/api-manual/video-series/bytedance-omnihuman-2-0-audio-to-video) | OmniHuman (ByteDance) | `bytedance/omnihuman-2.0/audio-to-video` |
| [PixVerse V6 Max Text to Video](/api-manual/video-series/pixverse-v6-max-text-to-video) | PixVerse | `pixverse/v6-max/text-to-video` |
| [PixVerse V6 Max Image to Video](/api-manual/video-series/pixverse-v6-max-image-to-video) | PixVerse | `pixverse/v6-max/image-to-video` |

## Notes

* Each model page lists the parameters, limits, and modes for its model IDs. Parameters are not shared across models.
* Generated files are valid for 24 hours after creation. Download and store them promptly.
* URL inputs must be publicly accessible HTTP(S) URLs. To use local files, upload them first with [File Management](/api-manual/file-series/upload-url).


# Veo 3.1
Source: https://docs.vidgo.ai/api-manual/video-series/veo-3-1

api-manual/video-series/veo-3-1.json POST /api/generate/submit
Fast and high-quality 8-second video generation with Google's VEO 3.1

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate fast or high-quality 8-second videos with Google's VEO 3.1 model. Supports text-to-video and image-to-video generation.

## Models

* **veo3.1-fast** - Fast 8-second generation
* **veo3.1-quality** - High-quality 8-second generation

## Features

* Asynchronous processing that returns a `task_id` for status tracking
* Supports text-to-video and image-to-video generation
* Supports up to 4k output resolution
* Generated video URLs are valid for 24 hours

## Duration Options

* **8 seconds** - Fixed duration for all generations

## Advanced Parameters

### Generation Type

* `frame` - Frame-to-video (two images)
* `reference` - Reference image video (three images)
* If omitted, inferred by `image_urls` count: 2 images for `frame`, 3 images for `reference`

### Image URLs

* Supports up to 3 images
* Frame mode: first image is the start frame, second image is the end frame
* Maximum file size: 10MB
* Supported formats: .jpeg, .jpg, .png, .webp

### Resolution

* `720p` (default), `1080p`, or `4k`


# Veo 3.1 Official
Source: https://docs.vidgo.ai/api-manual/video-series/veo-3-1-official

api-manual/video-series/veo-3-1-official.json POST /api/generate/submit
Official VEO 3.1 video generation with duration, audio control, and image-guided modes

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate videos with the official VEO 3.1 model family through one async API. The generation models support text-to-video, image-to-video, first/last-frame video, and three-image reference generation depending on the selected model and input.

## Models

* **veo3.1-fast-official** - Fast official generation with `4`, `6`, or `8` second duration options
* **veo3.1-lite-official** - Lightweight official generation with lower per-second pricing
* **veo3.1-quality-official** - Higher-quality official generation with 4K support

## Input modes

* No `image_urls`: text-to-video
* One image: image-to-video
* Two images: first/last-frame video. The first image is the start frame and the second image is the end frame
* Three images: reference generation. Use `generation_type: "reference"` and `duration: 8`
* `veo3.1-lite-official` supports at most two images and does not support `generation_type: "reference"`

## Parameters

* **model**: Required model identifier. Use `veo3.1-fast-official`, `veo3.1-lite-official`, or `veo3.1-quality-official`
* **prompt**: Required text prompt for video generation, up to `1000` characters
* **image\_urls**: Optional image URL array. Supports up to `3` public image URLs
* **generation\_type**: Optional `frame` or `reference`. If omitted, the mode is inferred from `image_urls`
* **duration**: `4`, `6`, or `8`. Default is `8`. Reference generation and `veo3.1-lite-official` with `resolution: "1080p"` support `8` only
* **aspect\_ratio**: `16:9` or `9:16`. `auto` is also supported for one-image and two-image workflows. Default is `16:9`
* **resolution**: `720p`, `1080p`, or `4k`. Default is `1080p`. `veo3.1-lite-official` does not support `4k`
* **sound**: Boolean audio switch. Default is `true`; set `false` for silent output
* **callback\_url**: Optional webhook URL for completion or failure notifications

## Pricing

Pricing is charged per generated second and varies by model, resolution, and whether audio is generated. See the model page for the current credit table.

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [google/veo3.1/text-to-video](/api-manual/video-series/google-veo3-1-text-to-video)
* [google/veo3.1/image-to-video](/api-manual/video-series/google-veo3-1-image-to-video)
* [google/veo3.1-fast/text-to-video](/api-manual/video-series/google-veo3-1-fast-text-to-video)
* [google/veo3.1-fast/image-to-video](/api-manual/video-series/google-veo3-1-fast-image-to-video)
* [google/veo3.1-lite/text-to-video](/api-manual/video-series/google-veo3-1-lite-text-to-video)
* [google/veo3.1-lite/image-to-video](/api-manual/video-series/google-veo3-1-lite-image-to-video)


# Veo 3.1 Fast Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/veo3-1-fast-image-to-video

api-manual/video-series/veo3-1-fast-image-to-video.json POST /api/generate/submit
Google Veo 3.1 Fast image to video model with fixed 8-second output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate an 8-second video with the Google Veo 3.1 Fast model.

## Models

* **veo3.1-fast/image-to-video** - Veo 3.1 Fast image to video generation

## Features

* Generate between an explicit starting frame and ending frame.
* Supports up to `4k` output resolution.
* Generated video URLs are temporary and should be downloaded promptly.

## Duration Options

* **8 seconds** - Fixed duration for this model.

## Advanced Parameters

### Generation Type

* The model ID fixes generation to `image-to-video`.

### Image URLs

* Provide exactly two images in starting-frame, ending-frame order.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `720p` (default), `1080p`, or `4k`.

## Pricing

Pricing varies by model variant and resolution. Check the Vidgo dashboard for the current rate.


# Veo 3.1 Fast Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/veo3-1-fast-reference-to-video

api-manual/video-series/veo3-1-fast-reference-to-video.json POST /api/generate/submit
Google Veo 3.1 Fast reference to video model with fixed 8-second output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate an 8-second video with the Google Veo 3.1 Fast model.

## Models

* **veo3.1-fast/reference-to-video** - Veo 3.1 Fast reference to video generation

## Features

* Guide Veo 3.1 Fast with exactly three reference images.
* Supports up to `4k` output resolution.
* Generated video URLs are temporary and should be downloaded promptly.

## Duration Options

* **8 seconds** - Fixed duration for this model.

## Advanced Parameters

### Generation Type

* The model ID fixes generation to `reference-to-video`.

### Image URLs

* Provide exactly three reference images.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `720p` (default), `1080p`, or `4k`.

## Pricing

Pricing varies by model variant and resolution. Check the Vidgo dashboard for the current rate.


# Veo 3.1 Fast Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/veo3-1-fast-text-to-video

api-manual/video-series/veo3-1-fast-text-to-video.json POST /api/generate/submit
Google Veo 3.1 Fast text to video model with fixed 8-second output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate an 8-second video with the Google Veo 3.1 Fast model.

## Models

* **veo3.1-fast/text-to-video** - Veo 3.1 Fast text to video generation

## Features

* Generate directly from a text prompt without image input.
* Supports up to `4k` output resolution.
* Generated video URLs are temporary and should be downloaded promptly.

## Duration Options

* **8 seconds** - Fixed duration for this model.

## Advanced Parameters

### Generation Type

* The model ID fixes generation to `text-to-video`.

### Resolution

* `720p` (default), `1080p`, or `4k`.

## Pricing

Pricing varies by model variant and resolution. Check the Vidgo dashboard for the current rate.


# Veo 3.1 Quality Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/veo3-1-image-to-video

api-manual/video-series/veo3-1-image-to-video.json POST /api/generate/submit
Google Veo 3.1 Quality image to video model with fixed 8-second output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate an 8-second video with the Google Veo 3.1 Quality model.

## Models

* **veo3.1/image-to-video** - Veo 3.1 Quality image to video generation

## Features

* Generate between an explicit starting frame and ending frame.
* Supports up to `4k` output resolution.
* Generated video URLs are temporary and should be downloaded promptly.

## Duration Options

* **8 seconds** - Fixed duration for this model.

## Advanced Parameters

### Generation Type

* The model ID fixes generation to `image-to-video`.

### Image URLs

* Provide exactly two images in starting-frame, ending-frame order.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `720p` (default), `1080p`, or `4k`.

## Pricing

Pricing varies by model variant and resolution. Check the Vidgo dashboard for the current rate.


# Veo 3.1 Lite Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/veo3-1-lite-text-to-video

api-manual/video-series/veo3-1-lite-text-to-video.json POST /api/generate/submit
Google Veo 3.1 Lite text to video model with fixed 8-second output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate an 8-second video with the Google Veo 3.1 Lite model.

## Models

* **veo3.1-lite/text-to-video** - Veo 3.1 Lite text to video generation

## Features

* Generate directly from a text prompt without image input.
* Supports up to `4k` output resolution.
* Generated video URLs are temporary and should be downloaded promptly.

## Duration Options

* **8 seconds** - Fixed duration for this model.

## Advanced Parameters

### Generation Type

* The model ID fixes generation to `text-to-video`.

### Resolution

* `720p` (default), `1080p`, or `4k`.
* `4k` is available on this model ID only. The `google/veo3.1-lite/text-to-video`, `google/veo3.1-lite/image-to-video` and `veo3.1-lite-official` IDs run a different Veo 3.1 Lite route and support `720p` and `1080p` only.

## Pricing

Pricing varies by model variant and resolution. Check the Vidgo dashboard for the current rate.


# Veo 3.1 Quality Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/veo3-1-text-to-video

api-manual/video-series/veo3-1-text-to-video.json POST /api/generate/submit
Google Veo 3.1 Quality text to video model with fixed 8-second output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate an 8-second video with the Google Veo 3.1 Quality model.

## Models

* **veo3.1/text-to-video** - Veo 3.1 Quality text to video generation

## Features

* Generate directly from a text prompt without image input.
* Supports up to `4k` output resolution.
* Generated video URLs are temporary and should be downloaded promptly.

## Duration Options

* **8 seconds** - Fixed duration for this model.

## Advanced Parameters

### Generation Type

* The model ID fixes generation to `text-to-video`.

### Resolution

* `720p` (default), `1080p`, or `4k`.

## Pricing

Pricing varies by model variant and resolution. Check the Vidgo dashboard for the current rate.


# Additional Models and Generated Files
Source: https://docs.vidgo.ai/api-manual/workflow-guide

Protocol support for additional chat models and file handling for audio and 3D workflows

## Chat model protocol support

Use `https://api.vidgo.ai` and send your API key as `Authorization: Bearer VIDGO_API_KEY`. Claude clients may also use `x-api-key`; Gemini clients may use `x-goog-api-key`.

Select the protocol supported by your model:

| Model | Chat Completions | Responses | Claude Messages | Gemini native |
| - | :-: | :-: | :-: | :-: |
| `openai/gpt-6-astra` | Yes | Yes | — | — |
| `anthropic/claude-fable-5.1` | Yes | — | Yes | — |
| `xai/grok-4.6` | Yes | Yes | — | — |
| `google/gemini-3.7-flash` | Yes | — | — | Yes |
| `google/gemini-3.5-flash` | Yes | — | — | Yes |
| `anthropic/claude-opus-4.8` | Yes | — | Yes | — |
| `openai/gpt-5.5` | Yes | Yes | — | — |
| `openai/gpt-5.4` | Yes | Yes | — | — |
| `openai/gpt-5.2` | Yes | Yes | — | — |
| `anthropic/claude-sonnet-4.6` | Yes | — | Yes | — |
| `anthropic/claude-opus-4.6` | Yes | — | Yes | — |

Both normal JSON responses and SSE streaming are supported. Use the exact model ID from the table with the request format for the corresponding endpoint:

* [Chat Completions](/api-manual/chat-series/chat-completions): `POST /v1/chat/completions`
* [Responses](/api-manual/chat-series/responses): `POST /v1/responses`
* [Claude Messages](/api-manual/chat-series/claude-messages): `POST /v1/messages`
* [Gemini Native Format](/api-manual/chat-series/gemini-native-format): `POST /v1beta/models/{model}:generateContent`

### Example request

```bash theme={null}
curl https://api.vidgo.ai/v1/chat/completions \
  -H "Authorization: Bearer $VIDGO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"openai/gpt-6-astra","messages":[{"role":"user","content":"Hello"}]}'
```

### Gemini model URLs

Use the complete public ID in the URL, including the provider slash. For example: `/v1beta/models/google/gemini-3.7-flash:generateContent`. Streaming uses `:streamGenerateContent?alt=sse`.

## Audio and 3D generated files

Audio and 3D workflows return a `task_id` through the unified submit API. Retrieve the completed results with [Query Task Status](/api-manual/task-management/status) or [Webhook Callbacks](/api-manual/task-management/webhooks).

Process every item in the returned `files` array. Depending on the workflow, results can include audio, model assets, previews, textures, animation files, lyrics, and timestamp files.

| Field | Type | Description |
| - | - | - |
| `file_url` | string | URL of the returned file. |
| `file_type` | string | File category: `image`, `video`, `audio`, `lyrics`, `3d`, or `other`. |
| `label` | string | Optional label identifying the file. |
| `format` | string | Optional file format. |
| `content_type` | string | Optional content type. |
| `file_name` | string | Optional file name. |
| `file_size` | integer | Optional file size in bytes. |
| `metadata` | object | Optional file metadata. |

Metadata fields are optional. Retain them when present, and do not assume that all files have the same type or that the first file is the only result.


# FAQs
Source: https://docs.vidgo.ai/faqs

Frequently asked questions about Vidgo API

Everything you need to know before you ship.

> [accordion] What is Vidgo API?
>
> Vidgo API is an AI infrastructure platform. It gives you unified access to top AI models with one key, predictable pricing, and production-ready reliability.

> [accordion] How do I get an API key?
>
> Create an account and generate your API key instantly from the dashboard.

> [accordion] Is there a free tier?
>
> New accounts receive free credits so you can test models before scaling.

> [accordion] Which models are available?
>
> We support leading image, video, and music models, with new additions every week.

> [accordion] Do you offer enterprise support?
>
> Yes. Contact our team to discuss custom pricing, SLAs, and dedicated support.

> [accordion] Where can I check current rates?
>
> See the [Pricing page](https://vidgo.ai/apis/api-pricing) or your Vidgo dashboard for current model rates.

> [accordion] What affects generation costs?
>
> Cost depends on the selected model and its supported billing options. Image generation can depend on quality, resolution, output count, and reference image count. Video generation can depend on duration, resolution, audio generation, and reference-video duration. See each model page for the factors that apply to that workflow.

> [accordion] Which music model do you use?
>
> We use Suno's official models, including the latest v5 model. Our music API supports generating tracks up to 8 minutes long with professional quality.

> [accordion] Will I be charged for failed requests?
>
> No. We have transparent pricing and only charge for successful generations. Failed requests will automatically refund your credits.

> [accordion] What payment methods do you support?
>
> We support multiple payment methods including Stripe (credit/debit cards), WeChat Pay, and cryptocurrency payments for maximum flexibility.

> [accordion] How can I contact support if I have issues?
>
> You can reach us at [support@vidgo.ai](mailto:support@vidgo.ai). We're also gradually launching additional support channels like Discord - stay tuned for announcements on our website.


# Getting Started
Source: https://docs.vidgo.ai/index

Make your first Vidgo API request and learn the request lifecycle for image, video, audio, music, 3D, and LLM models.

# Build with Vidgo API

Access image, video, audio, music, 3D, and LLM models through **one API key**. Generation models share one asynchronous submit-and-query workflow; LLM models use OpenAI-, Claude-, and Gemini-compatible endpoints.

> [card] [Get an API key](https://vidgo.ai/apis/dashboard/api-key)
>
> Create and manage your keys in the Vidgo dashboard.

> [card] [Make your first request](/api-manual/quickstart)
>
> Submit a task and poll its result with cURL, Python, or JavaScript.

> [card] [Check pricing](https://vidgo.ai/apis/api-pricing)
>
> See current model rates.

> [card] [Handle errors](/api-manual/error-codes)
>
> Response formats and HTTP status codes.

> [card] [Build with your AI agent](https://vidgo.ai/apis/agent-setup)
>
> Connect Claude Code, Codex, Cursor or OpenCode to the Vidgo docs MCP server (`https://docs.vidgo.ai/mcp`) and copy a setup prompt that works in any coding agent.

## 1. Available models

Each model category has an overview listing every model and its model IDs:

> [card] [Image](/api-manual/image-series/overview)
>
> Text-to-image and image editing.

> [card] [Video](/api-manual/video-series/overview)
>
> Text, image, and reference to video.

> [card] [Audio](/api-manual/music-series/audio-overview)
>
> Text-to-speech.

> [card] [LLM](/api-manual/chat-series/overview)
>
> GPT, Claude, Gemini, Grok, and Kimi.

> [card] [3D](/api-manual/3d-series/overview)
>
> Text and image to 3D models.

> [card] [Music](/api-manual/music-series/overview)
>
> Suno music actions and music models.

## 2. Authentication

Create your API key on the [API key page](https://vidgo.ai/apis/dashboard/api-key). Every request must include:

```http theme={"dark"}
Authorization: Bearer VIDGO_API_KEY
Content-Type: application/json
```

[Claude Messages](/api-manual/chat-series/claude-messages) also accepts `x-api-key`, and [Gemini Native Format](/api-manual/chat-series/gemini-native-format) also accepts `x-goog-api-key`.

> [!WARNING]
>
> **Never expose your API key** in frontend code (browser, mobile apps, public repositories). Keep it on your server.

## 3. Submit a generation task

Generation is asynchronous. Submit a task to `POST /api/generate/submit` and save the returned `task_id`.

```bash theme={"dark"}
curl https://api.vidgo.ai/api/generate/submit \
  -H "Authorization: Bearer $VIDGO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-image",
    "input": {
      "prompt": "A paper-cut illustration of a city at sunrise",
      "size": "1:1"
    }
  }'
```

> [!NOTE]
>
> A successful submit response means the task was accepted, not that generation has finished.

## 4. Retrieve the result

```bash theme={"dark"}
curl "https://api.vidgo.ai/api/generate/status/TASK_ID" \
  -H "Authorization: Bearer $VIDGO_API_KEY"
```

**1. Submit**

Send the generation request and save the returned `task_id`.

**2. Wait for the result**

Poll [Get Task Status](/api-manual/task-management/status), or pass a `callback_url` and receive a [webhook](/api-manual/task-management/webhooks) when the task reaches `finished` or `failed`.

**3. Store the output**

Download the files in `files` and store them on your side.

Suno music actions use [Query Music Detail](/api-manual/music-series/query-music-detail) and the [Music Webhook](/api-manual/music-series/music-webhook) instead of the unified task status endpoint.

## 5. Data retention

> [!WARNING]
>
> Generated files are valid for **24 hours**. Download and store results promptly.

## 6. Before production

> [!TIP]
>
> Keep API keys on your server; never expose them in browser or mobile code.

> [!TIP]
>
> Validate inputs against the selected model page before submitting.

> [!TIP]
>
> Handle `400`, `401`, `429`, and transient `5xx` responses explicitly; see [Error Codes](/api-manual/error-codes).

> [!TIP]
>
> Treat every submission retry as potentially new paid work; confirm whether the original request created a task before resubmitting.

> [!TIP]
>
> Check your remaining credits with [Query User Balance](/api-manual/account-management/user-balance).

## 7. Support

> [card] [Email](mailto:support@vidgo.ai)
>
> `support@vidgo.ai`

> [card] [FAQs](/faqs)
>
> Billing, models, and common questions.


# Happy Horse Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-happyhorse-1-0-image-to-video

api-manual/video-series/alibaba-happyhorse-1-0-image-to-video.json POST /api/generate/submit
Animate a still image with Happy Horse, using text to guide the scene and motion in videos up to fifteen seconds long.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate a still image with Happy Horse, using text to guide the scene and motion in videos up to fifteen seconds long.

## Models

* **alibaba/happyhorse-1.0/image-to-video** - Generate video from image input.

## Features

* Animate a single reference image.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **2,500 characters**.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `1080p`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Happy Horse Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-happyhorse-1-0-reference-to-video

api-manual/video-series/alibaba-happyhorse-1-0-reference-to-video.json POST /api/generate/submit
Use reference images to shape new Happy Horse video scenes, combining visual direction with a written description at up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Use reference images to shape new Happy Horse video scenes, combining visual direction with a written description at up to 1080p.

## Models

* **alibaba/happyhorse-1.0/reference-to-video** - Generate a new video guided by reference material and a prompt.

## Features

* Guide video generation with one to nine reference images.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **2,500 characters**.

### Reference Inputs

* `reference_image_urls`: Required; provide up to 9 images.
* Describe referenced subjects in the prompt; `character1` and `character2` markers are not required.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `1080p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`. Default: `16:9`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Happy Horse Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-happyhorse-1-0-text-to-video

api-manual/video-series/alibaba-happyhorse-1-0-text-to-video.json POST /api/generate/submit
Create short scenes from text with Happy Horse, with output up to 1080p for landscape, portrait, and square video.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create short scenes from text with Happy Horse, with output up to 1080p for landscape, portrait, and square video.

## Models

* **alibaba/happyhorse-1.0/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **2,500 characters**.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `1080p`.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`. Default: `16:9`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Happy Horse Video Edit
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-happyhorse-1-0-video-edit

api-manual/video-series/alibaba-happyhorse-1-0-video-edit.json POST /api/generate/submit
Transform existing footage with Happy Horse using text instructions and reference images to guide a new visual direction.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Transform existing footage with Happy Horse using text instructions and reference images to guide a new visual direction.

## Models

* **alibaba/happyhorse-1.0/video-edit** - Edit an existing video using written instructions.

## Features

* Edit an existing video using written instructions.
* Optionally add up to five reference images to guide the edit.
* Choose from `720p`, `1080p` output resolutions.
* Choose automatic audio handling or preserve the source audio.

## Duration Options

* **Source video: 3-60 seconds** - Duration is read from `video_url`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the requested changes to the video.
* The prompt must not be blank. Maximum length: **2,500 characters**.

### Reference Inputs

* `reference_image_urls`: Optional; provide up to 5 images.
* Describe referenced subjects in the prompt; `character1` and `character2` markers are not required.

### Source Video

* `video_url`: Required. Provide one publicly accessible source video URL.
* The source video must be directly downloadable over HTTP(S) and no larger than **100 MiB** for duration inspection.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `1080p`.

### Audio

* `audio_setting`: `auto` or `origin`. Default: `auto`.
* `auto` lets the model determine audio handling; `origin` preserves the original audio.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking.

## Pricing

Credits are calculated from the detected source-video duration, capped at **15 seconds**, multiplied by the selected resolution rate.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Happy Horse 1.1 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-happyhorse-1-1-image-to-video

api-manual/video-series/alibaba-happyhorse-1-1-image-to-video.json POST /api/generate/submit
Bring a still image into motion with Happy Horse 1.1, creating video scenes up to fifteen seconds long in a range of screen formats.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Bring a still image into motion with Happy Horse 1.1, creating video scenes up to fifteen seconds long in a range of screen formats.

## Models

* **alibaba/happyhorse-1.1/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Accepts exactly 1 image in `image_urls`.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials. Image inputs also accept image Data URIs or raw Base64. Image conversion and media validation complete before credits are deducted.

`duration` and `seed` also accept integer strings such as `"5"`; decimal numbers such as `5.0` and decimal strings such as `"5.0"` are rejected. Null or empty values use the default duration or omit the seed. Supported boolean fields also accept `0`/`1` and case-insensitive strings `true`/`false`, `1`/`0`, `yes`/`no`, `y`/`n`, and `on`/`off`; null omits the field. Surrounding whitespace is trimmed, and resolution values are case-insensitive.

### Prompt

* `prompt`: Optional. Describe the scene and desired motion. Maximum: **2,500 Unicode characters**.
* Leading and trailing whitespace is trimmed. Omitted, null, empty, and whitespace-only values are treated as omitted.

### Image URLs

* `image_urls`: Required. Supply exactly 1 item. Each item may be a public HTTP(S) URL, image Data URI, or raw Base64. Image order is preserved.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `1080p`.

### Seed

* `seed`: Optional. Set the generation seed. Integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`.

## Pricing

Usage equals output seconds multiplied by the selected resolution rate. See [Pricing](https://vidgo.ai/apis/api-pricing) for current rates.

Credits are deducted on submission and refunded if the task fails.


# Happy Horse 1.1 Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-happyhorse-1-1-reference-to-video

api-manual/video-series/alibaba-happyhorse-1-1-reference-to-video.json POST /api/generate/submit
Guide Happy Horse 1.1 video generation with up to nine reference images to establish subjects, settings, and visual direction.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Guide Happy Horse 1.1 video generation with up to nine reference images to establish subjects, settings, and visual direction.

## Models

* **alibaba/happyhorse-1.1/reference-to-video** - Generate video guided by reference material and a prompt.

## Features

* Generate video guided by reference material and a prompt.
* Use reference images to guide the generated scene.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials. Image inputs also accept image Data URIs or raw Base64. Image conversion and media validation complete before credits are deducted.

`duration` and `seed` also accept integer strings such as `"5"`; decimal numbers such as `5.0` and decimal strings such as `"5.0"` are rejected. Null or empty values use the default duration or omit the seed. Supported boolean fields also accept `0`/`1` and case-insensitive strings `true`/`false`, `1`/`0`, `yes`/`no`, `y`/`n`, and `on`/`off`; null omits the field. Surrounding whitespace is trimmed, and resolution values are case-insensitive.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Maximum: **2,500 Unicode characters**.
* Leading and trailing whitespace is trimmed. A nonblank prompt is required.

### Reference Images

* `reference_image_urls`: Required. Supply 1-9 items. Each item may be a public HTTP(S) URL, image Data URI, or raw Base64. Image order is preserved.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, `4:5`, `5:4`, `9:16`, `9:21`. Default: `16:9`.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `1080p`.

### Seed

* `seed`: Optional. Set the generation seed. Integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`.

## Pricing

Usage equals output seconds multiplied by the selected resolution rate. See [Pricing](https://vidgo.ai/apis/api-pricing) for current rates.

Credits are deducted on submission and refunded if the task fails.


# Happy Horse 1.1 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-happyhorse-1-1-text-to-video

api-manual/video-series/alibaba-happyhorse-1-1-text-to-video.json POST /api/generate/submit
Generate short video scenes with Happy Horse 1.1, with output up to 1080p and framing from wide cinematic layouts to tall vertical video.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate short video scenes with Happy Horse 1.1, with output up to 1080p and framing from wide cinematic layouts to tall vertical video.

## Models

* **alibaba/happyhorse-1.1/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

`duration` and `seed` also accept integer strings such as `"5"`; decimal numbers such as `5.0` and decimal strings such as `"5.0"` are rejected. Null or empty values use the default duration or omit the seed. Supported boolean fields also accept `0`/`1` and case-insensitive strings `true`/`false`, `1`/`0`, `yes`/`no`, `y`/`n`, and `on`/`off`; null omits the field. Surrounding whitespace is trimmed, and resolution values are case-insensitive.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Maximum: **2,500 Unicode characters**.
* Leading and trailing whitespace is trimmed. A nonblank prompt is required.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, `4:5`, `5:4`, `9:16`, `9:21`. Default: `16:9`.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `1080p`.

### Seed

* `seed`: Optional. Set the generation seed. Integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`.

## Pricing

Usage equals output seconds multiplied by the selected resolution rate. See [Pricing](https://vidgo.ai/apis/api-pricing) for current rates.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.5 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-5-image-to-video

api-manual/video-series/alibaba-wan-2-5-image-to-video.json POST /api/generate/submit
Animate one reference image into a 5 or 10 second video with synchronized audio and output up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate one reference image into a 5 or 10 second video with synchronized audio and output up to 1080p.

## Models

* `alibaba/wan-2.5/image-to-video`: Animate a reference image using a motion prompt.

## Request Parameters

Set `model` and optional `callback_url` at the request root. Place the following generation parameters inside `input`.

### prompt

* `prompt` (string, required): Describe the subject, scene, and motion. Use 1–1500 characters after trimming surrounding whitespace.

### image\_urls

* `image_urls` (string\[], required): Provide exactly one publicly accessible HTTP(S) image URL.

### resolution

* `resolution` (string, optional): Output resolution: 480p, 720p, or 1080p. Default: 720p.

### duration

* `duration` (integer, optional): Video duration: 5 or 10 seconds. Default: 5.

### audio

* `audio` (string, optional): Optional audio file URL to supply a recording for the video.

### negative\_prompt

* `negative_prompt` (string, optional): Optional description of elements to exclude from the scene.

### seed

* `seed` (integer, optional): Optional integer random seed for reusing generation settings.

## Retrieve the Video

Query the task with `task_id`. Continue polling while `status` is `not_started` or `running`. At `finished`, read the video URL from `data.files[].file_url`. At `failed`, read `data.error_message` for the error details.


# Wan 2.5 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-5-text-to-video

api-manual/video-series/alibaba-wan-2-5-text-to-video.json POST /api/generate/submit
Create 5 or 10 second videos from text with synchronized audio and landscape or portrait output up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create 5 or 10 second videos from text with synchronized audio and landscape or portrait output up to 1080p.

## Models

* `alibaba/wan-2.5/text-to-video`: Generate a video scene from a text prompt.

## Request Parameters

Set `model` and optional `callback_url` at the request root. Place the following generation parameters inside `input`.

### prompt

* `prompt` (string, required): Describe the subject, scene, and motion. Use 1–1500 characters after trimming surrounding whitespace.

### aspect\_ratio

* `aspect_ratio` (string, required): Frame dimensions in pixels (`width*height`): `832*480`, `480*832`, `1280*720`, `720*1280`, `1920*1080`, `1080*1920`.

### duration

* `duration` (integer, optional): Video duration: 5 or 10 seconds. Default: 5.

### audio

* `audio` (string, optional): Optional audio file URL to supply a recording for the video.

### negative\_prompt

* `negative_prompt` (string, optional): Optional description of elements to exclude from the scene.

### seed

* `seed` (integer, optional): Optional integer random seed for reusing generation settings.

## Retrieve the Video

Query the task with `task_id`. Continue polling while `status` is `not_started` or `running`. At `finished`, read the video URL from `data.files[].file_url`. At `failed`, read `data.error_message` for the error details.


# Wan 2.6 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-6-image-to-video

api-manual/video-series/alibaba-wan-2-6-image-to-video.json POST /api/generate/submit
Bring a still image into motion with Wan 2.6, creating video sequences with optional multiple shots at up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Bring a still image into motion with Wan 2.6, creating video sequences with optional multiple shots at up to 1080p.

## Models

* **alibaba/wan-2.6/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Accepts exactly 1 image in `image_urls`.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **5, 10, 15 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.
* The playground accepts JPG, PNG, and WebP uploads up to 10 MiB.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.6 Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-6-reference-to-video

api-manual/video-series/alibaba-wan-2-6-reference-to-video.json POST /api/generate/submit
Use reference videos to guide new scenes with Wan 2.6, with support for multiple shots and video output up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Use reference videos to guide new scenes with Wan 2.6, with support for multiple shots and video output up to 1080p.

## Models

* **alibaba/wan-2.6/reference-to-video** - Generate video guided by reference material and a prompt.

## Features

* Generate video guided by reference material and a prompt.
* Use 1-10 video references to guide the generated scene.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **5, 10 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Reference Videos

* `video_urls`: Required. Provide reference video URLs. Supply 1-10 items. Each item must be an HTTP(S) URL.
* Reference videos support MP4, MOV, and MKV files up to 10 MiB each.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.6 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-6-text-to-video

api-manual/video-series/alibaba-wan-2-6-text-to-video.json POST /api/generate/submit
Turn written scenes into video with Wan 2.6, with support for multiple shots and output up to fifteen seconds at 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn written scenes into video with Wan 2.6, with support for multiple shots and output up to fifteen seconds at 1080p.

## Models

* **alibaba/wan-2.6/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **5, 10, 15 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.7 Video Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-7-image-to-video

api-manual/video-series/alibaba-wan-2-7-image-to-video.json POST /api/generate/submit
Animate reference images with Wan 2.7, with optional audio guidance and multiple shots for video sequences up to fifteen seconds.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate reference images with Wan 2.7, with optional audio guidance and multiple shots for video sequences up to fifteen seconds.

## Models

* **alibaba/wan-2.7/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Accepts 1-2 images in `image_urls`.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **2-15 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials. Image inputs also accept image Data URIs or raw Base64. Image conversion and media validation complete before credits are deducted.

`duration` and `seed` also accept integer strings such as `"5"`; decimal numbers such as `5.0` and decimal strings such as `"5.0"` are rejected. Null or empty values use the default duration or omit the seed. Supported boolean fields also accept `0`/`1` and case-insensitive strings `true`/`false`, `1`/`0`, `yes`/`no`, `y`/`n`, and `on`/`off`; null omits the field. Surrounding whitespace is trimmed, and resolution values are case-insensitive.

### Prompt

* `prompt`: Optional. Describe the scene and desired motion. Maximum: **5,000 Unicode characters**.
* Leading and trailing whitespace is trimmed. Omitted, null, empty, and whitespace-only values are treated as omitted.

### Image URLs

* `image_urls`: Required. Supply 1-2 items. Each item may be a public HTTP(S) URL, image Data URI, or raw Base64. Image order is preserved.

### Audio URL

* `audio_url`: Optional. Provide the input audio URL. Use a publicly accessible HTTP(S) URL.

### Video URL

* `video_url`: Optional. Provide a publicly accessible HTTP(S) video URL.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`.

### Seed

* `seed`: Optional. Set the generation seed. Integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`.

## Pricing

Usage equals output seconds multiplied by the selected resolution rate. See [Pricing](https://vidgo.ai/apis/api-pricing) for current rates.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.7 Video Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-7-reference-to-video

api-manual/video-series/alibaba-wan-2-7-reference-to-video.json POST /api/generate/submit
Combine image and video references to guide Wan 2.7 scenes, with support for multiple shots and output up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Combine image and video references to guide Wan 2.7 scenes, with support for multiple shots and output up to 1080p.

## Models

* **alibaba/wan-2.7/reference-to-video** - Generate video guided by reference material and a prompt.

## Features

* Generate video guided by reference material and a prompt.
* Use reference images to guide the generated scene.
* Use video references to guide the generated scene.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **2-10 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials. Image inputs also accept image Data URIs or raw Base64. Image conversion and media validation complete before credits are deducted.

* Provide at least one nonempty reference array: `reference_image_urls` or `reference_video_urls`.

`duration` and `seed` also accept integer strings such as `"5"`; decimal numbers such as `5.0` and decimal strings such as `"5.0"` are rejected. Null or empty values use the default duration or omit the seed. Supported boolean fields also accept `0`/`1` and case-insensitive strings `true`/`false`, `1`/`0`, `yes`/`no`, `y`/`n`, and `on`/`off`; null omits the field. Surrounding whitespace is trimmed, and resolution values are case-insensitive.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Maximum: **5,000 Unicode characters**.
* Leading and trailing whitespace is trimmed. A nonblank prompt is required.

### Reference Images

* `reference_image_urls`: Optional. Supply up to 9 items. Each item may be a public HTTP(S) URL, image Data URI, or raw Base64. Image order is preserved.

### Reference Videos

* `reference_video_urls`: Optional. Provide reference video URLs. Supply an array. Each item must be an HTTP(S) URL.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`. Default: `16:9`.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`.

### Seed

* `seed`: Optional. Set the generation seed. Integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`.

## Pricing

Usage equals output seconds multiplied by the selected resolution rate. See [Pricing](https://vidgo.ai/apis/api-pricing) for current rates.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.7 Video Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-7-text-to-video

api-manual/video-series/alibaba-wan-2-7-text-to-video.json POST /api/generate/submit
Create 5, 10, or 15-second videos with Wan 2.7 using written scenes and optional audio input at up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create 5, 10, or 15-second videos with Wan 2.7 using written scenes and optional audio input at up to 1080p.

## Models

* **alibaba/wan-2.7/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **5, 10, 15 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

`duration` and `seed` also accept integer strings such as `"5"`; decimal numbers such as `5.0` and decimal strings such as `"5.0"` are rejected. Null or empty values use the default duration or omit the seed. Supported boolean fields also accept `0`/`1` and case-insensitive strings `true`/`false`, `1`/`0`, `yes`/`no`, `y`/`n`, and `on`/`off`; null omits the field. Surrounding whitespace is trimmed, and resolution values are case-insensitive.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Maximum: **5,000 Unicode characters**.
* Leading and trailing whitespace is trimmed. A nonblank prompt is required.

### Audio URL

* `audio_url`: Optional. Provide the input audio URL. Use a publicly accessible HTTP(S) URL.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`. Default: `16:9`.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`.

### Seed

* `seed`: Optional. Set the generation seed. Integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`.

## Pricing

Usage equals output seconds multiplied by the selected resolution rate. See [Pricing](https://vidgo.ai/apis/api-pricing) for current rates.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.7 Video Edit
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-2-7-video-edit

api-manual/video-series/alibaba-wan-2-7-video-edit.json POST /api/generate/submit
Rework existing footage with Wan 2.7 using text and an optional reference image to guide changes to the scene.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Rework existing footage with Wan 2.7 using text and an optional reference image to guide changes to the scene.

## Models

* **alibaba/wan-2.7/video-edit** - Edit an existing video using written instructions.

## Features

* Edit an existing video using written instructions.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **Automatic** - `duration: 0` is the default. The service detects the source-video duration, checks that it is **2-10 seconds**, then rounds down to whole seconds. The same whole-second duration is used for billing and output generation.
* **Explicit output length** - Set `duration` to `2`, `3`, `4`, `5`, `6`, `7`, `8`, `9`, `10` seconds.
* With `duration: 0`, the source must be a publicly reachable HTTP(S) file no larger than **100 MiB**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials. Image inputs also accept image Data URIs or raw Base64. Image conversion and media validation complete before credits are deducted.

`duration` and `seed` also accept integer strings such as `"5"`; decimal numbers such as `5.0` and decimal strings such as `"5.0"` are rejected. Null or empty values use the default duration or omit the seed. Supported boolean fields also accept `0`/`1` and case-insensitive strings `true`/`false`, `1`/`0`, `yes`/`no`, `y`/`n`, and `on`/`off`; null omits the field. Surrounding whitespace is trimmed, and resolution values are case-insensitive.

### Prompt

* `prompt`: Required. Describe the requested changes. Maximum: **5,000 Unicode characters**.
* Leading and trailing whitespace is trimmed. A nonblank prompt is required.

### Reference Image URL

* `reference_image_url`: Optional. Provide an image to guide the edit as a public HTTP(S) URL, image Data URI, or raw Base64.

### Source Video

* `video_url`: Required. Provide the source video URL. Use a publicly accessible HTTP(S) URL.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `16:9`, `9:16`, `1:1`, `4:3`, `3:4`. Default: `16:9`.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`.

### Audio Settings

* `audio_setting`: Optional. Supply the audio setting as a string.

### Seed

* `seed`: Optional. Set the generation seed. Integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`.

## Pricing

Usage equals output seconds multiplied by the selected resolution rate. See [Pricing](https://vidgo.ai/apis/api-pricing) for current rates.

Credits are deducted on submission and refunded if the task fails.


# Wan 3.0 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-3-0-image-to-video

api-manual/video-series/alibaba-wan-3-0-image-to-video.json POST /api/generate/submit
Animate still images with Wan 3.0, combining text direction with optional audio in clips up to thirty seconds long.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate still images with Wan 3.0, combining text direction with optional audio in clips up to thirty seconds long.

## Models

* **alibaba/wan-3.0/image-to-video** - Generate video from image input.

## Features

* Generate video from one or two input images.
* Choose from `480p`, `720p`, `1080p` output resolutions.
* Control audio generation with `audio`; it defaults to `true`.

## Duration Options

* **2-30 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Optional. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Image URLs

* `image_urls`: Required. Provide 1-2 image URLs.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `resolution`: `480p`, `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `adaptive`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `adaptive`.

### Audio

* `audio`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan 3.0 Video Prime Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-3-0-prime-image-to-video

api-manual/video-series/alibaba-wan-3-0-prime-image-to-video.json POST /api/generate/submit
Animate still images with Wan 3.0 Prime, combining text direction with optional audio in clips up to thirty seconds long.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate still images with Wan 3.0 Prime, combining text direction with optional audio in clips up to thirty seconds long.

## Models

* **alibaba/wan-3.0-prime/image-to-video** - Generate video from image input.

## Features

* Generate video from one or two input images.
* Choose from `480p`, `720p`, `1080p` output resolutions.
* Control audio generation with `audio`; it defaults to `true`.

## Duration Options

* **2-30 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Optional. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Image URLs

* `image_urls`: Required. Provide 1-2 image URLs.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `resolution`: `480p`, `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `adaptive`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `adaptive`.

### Audio

* `audio`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan 3.0 Video Prime Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-3-0-prime-reference-to-video

api-manual/video-series/alibaba-wan-3-0-prime-reference-to-video.json POST /api/generate/submit
Guide Wan 3.0 Prime video scenes with image, video, and audio references, combining multiple sources into a new visual sequence.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Guide Wan 3.0 Prime video scenes with image, video, and audio references, combining multiple sources into a new visual sequence.

## Models

* **alibaba/wan-3.0-prime/reference-to-video** - Generate a new video guided by reference material and a prompt.

## Features

* Use image, video, audio, document, or webpage references to guide a new scene.
* Choose from `480p`, `720p`, `1080p` output resolutions.
* Control audio generation with `audio`; it defaults to `true`.

## Duration Options

* **2-30 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Reference Inputs

* At least one reference type is required in addition to `prompt`. Audio can be the only reference type.
* Do not combine `reference_file_urls` with `reference_link_urls`.
* `reference_image_urls`: Optional; provide 1-10 images.
* `reference_video_urls`: Optional; provide 1-5 videos.
* `reference_audio_urls`: Optional; provide 1-5 audio files.
* `reference_file_urls`: Optional; provide exactly 1 document.
* `reference_link_urls`: Optional; provide exactly 1 webpage.

### Resolution

* `resolution`: `480p`, `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `adaptive`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `adaptive`.

### Audio

* `audio`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan 3.0 Video Prime Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-3-0-prime-text-to-video

api-manual/video-series/alibaba-wan-3-0-prime-text-to-video.json POST /api/generate/submit
Create video scenes with Wan 3.0 Prime, with optional audio and output up to thirty seconds for longer visual sequences.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create video scenes with Wan 3.0 Prime, with optional audio and output up to thirty seconds for longer visual sequences.

## Models

* **alibaba/wan-3.0-prime/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `480p`, `720p`, `1080p` output resolutions.
* Control audio generation with `audio`; it defaults to `true`.

## Duration Options

* **2-30 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Resolution

* `resolution`: `480p`, `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `adaptive`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `adaptive`.

### Audio

* `audio`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan 3.0 Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-3-0-reference-to-video

api-manual/video-series/alibaba-wan-3-0-reference-to-video.json POST /api/generate/submit
Guide Wan 3.0 video scenes with image, video, and audio references, combining multiple sources into a new visual sequence.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Guide Wan 3.0 video scenes with image, video, and audio references, combining multiple sources into a new visual sequence.

## Models

* **alibaba/wan-3.0/reference-to-video** - Generate a new video guided by reference material and a prompt.

## Features

* Use image, video, audio, document, or webpage references to guide a new scene.
* Choose from `480p`, `720p`, `1080p` output resolutions.
* Control audio generation with `audio`; it defaults to `true`.

## Duration Options

* **2-30 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Reference Inputs

* At least one reference type is required in addition to `prompt`. Audio can be the only reference type.
* Do not combine `reference_file_urls` with `reference_link_urls`.
* `reference_image_urls`: Optional; provide 1-10 images.
* `reference_video_urls`: Optional; provide 1-5 videos.
* `reference_audio_urls`: Optional; provide 1-5 audio files.
* `reference_file_urls`: Optional; provide exactly 1 document.
* `reference_link_urls`: Optional; provide exactly 1 webpage.

### Resolution

* `resolution`: `480p`, `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `adaptive`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `adaptive`.

### Audio

* `audio`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan 3.0 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-3-0-text-to-video

api-manual/video-series/alibaba-wan-3-0-text-to-video.json POST /api/generate/submit
Create video scenes with Wan 3.0, with optional audio and output up to thirty seconds for longer visual sequences.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create video scenes with Wan 3.0, with optional audio and output up to thirty seconds for longer visual sequences.

## Models

* **alibaba/wan-3.0/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `480p`, `720p`, `1080p` output resolutions.
* Control audio generation with `audio`; it defaults to `true`.

## Duration Options

* **2-30 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied. Maximum length: **20,000 characters**.

### Resolution

* `resolution`: `480p`, `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `adaptive`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `adaptive`.

### Audio

* `audio`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

### Seed

* `seed` is optional. Supply an integer from `0` to `2147483647`.

### Safety Checker

* `enable_safety_checker` is optional. Set it to `true` or `false` to control safety checking. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Wan Animate Move
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-v2-2-14b-animate-move

api-manual/video-series/alibaba-wan-v2-2-14b-animate-move.json POST /api/generate/submit
Animate the character in an image using motion from a reference video with Wan Animate Move.

> [!TIP]
>
> 1. Submit the request with `VIDGO_API_KEY`. Vidgo validates the request and checks the source video before creating the task. A `task_id` is returned after the task is successfully created.
> 2. Use the returned `task_id` with [Query Task Status](/api-manual/task-management/status) until the task reaches `finished` or `failed`. If you provide a `callback_url`, Vidgo also sends the result to your [webhook](/api-manual/task-management/webhooks) when the task finishes or fails.

Animate the character in an image using motion from a reference video with Wan Animate Move.

## Models

* **alibaba/wan/v2.2-14b/animate/move** - Animate the character in an image using motion from a reference video with Wan Animate Move.

## Features

* Animate the character in an image using motion from a reference video with Wan Animate Move.
* Accepts an `image_urls` array containing exactly one image URL.
* Supports `480p`, `580p`, and `720p` output resolutions.

## Source Video Requirements

* `video_url` must point to a publicly accessible HTTP(S) video file.
* The entire video file must be no larger than **100 MiB** (104,857,600 bytes), and its duration must be greater than zero.
* Vidgo reads the source video to determine its duration for billing.

## Request Parameters

Set `model` and optional `callback_url` at the request root. The `input` object accepts only `video_url`, `image_urls`, and `resolution`.

Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials or whitespace.

### Source Video

* `video_url`: Required string. The reference video that provides the motion for the character image. The video must meet the source video requirements above.

### Character Image

* `image_urls`: Required array containing exactly one publicly accessible HTTP(S) image URL. The image containing the character to animate. A single URL string instead of an array, an empty array, multiple URLs, and Base64 images are rejected.

### Resolution

* `resolution`: Optional string. Accepted values: `480p`, `580p`, `720p`. Defaults to `480p` only when the field is omitted. Null, empty strings, other values, and values with surrounding whitespace are rejected.

## Billing Factors

Billing depends on the selected resolution (`resolution`) and the measured source-video duration (`video_url`). Move and Replace use the same rates.

Source-video duration is rounded up to whole seconds before billing. For example, a source video lasting 5.1 seconds is billed as 6 seconds. If Vidgo cannot read the video or determine its duration, the request is rejected before credits are deducted.

Credits are deducted when a valid request is submitted for generation and refunded if the task fails.


# Wan Animate Replace
Source: https://docs.vidgo.ai/api-manual/video-series/alibaba-wan-v2-2-14b-animate-replace

api-manual/video-series/alibaba-wan-v2-2-14b-animate-replace.json POST /api/generate/submit
Replace the character in a source video with the character from a reference image using Wan Animate Replace.

> [!TIP]
>
> 1. Submit the request with `VIDGO_API_KEY`. Vidgo validates the request and checks the source video before creating the task. A `task_id` is returned after the task is successfully created.
> 2. Use the returned `task_id` with [Query Task Status](/api-manual/task-management/status) until the task reaches `finished` or `failed`. If you provide a `callback_url`, Vidgo also sends the result to your [webhook](/api-manual/task-management/webhooks) when the task finishes or fails.

Replace the character in a source video with the character from a reference image using Wan Animate Replace.

## Models

* **alibaba/wan/v2.2-14b/animate/replace** - Replace the character in a source video with the character from a reference image using Wan Animate Replace.

## Features

* Replace the character in a source video with the character from a reference image using Wan Animate Replace.
* Accepts an `image_urls` array containing exactly one image URL.
* Supports `480p`, `580p`, and `720p` output resolutions.

## Source Video Requirements

* `video_url` must point to a publicly accessible HTTP(S) video file.
* The entire video file must be no larger than **100 MiB** (104,857,600 bytes), and its duration must be greater than zero.
* Vidgo reads the source video to determine its duration for billing.

## Request Parameters

Set `model` and optional `callback_url` at the request root. The `input` object accepts only `video_url`, `image_urls`, and `resolution`.

Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials or whitespace.

### Source Video

* `video_url`: Required string. The source video containing the character to replace. The video must meet the source video requirements above.

### Character Image

* `image_urls`: Required array containing exactly one publicly accessible HTTP(S) image URL. The reference image containing the replacement character. A single URL string instead of an array, an empty array, multiple URLs, and Base64 images are rejected.

### Resolution

* `resolution`: Optional string. Accepted values: `480p`, `580p`, `720p`. Defaults to `480p` only when the field is omitted. Null, empty strings, other values, and values with surrounding whitespace are rejected.

## Billing Factors

Billing depends on the selected resolution (`resolution`) and the measured source-video duration (`video_url`). Move and Replace use the same rates.

Source-video duration is rounded up to whole seconds before billing. For example, a source video lasting 5.1 seconds is billed as 6 seconds. If Vidgo cannot read the video or determine its duration, the request is rejected before credits are deducted.

Credits are deducted when a valid request is submitted for generation and refunded if the task fails.


# FLUX 3 Extend Video
Source: https://docs.vidgo.ai/api-manual/video-series/blackforestlabs-flux-3-extend-video

api-manual/video-series/blackforestlabs-flux-3-extend-video.json POST /api/generate/submit
Continue an existing video with FLUX 3, using a text prompt to guide the next part of the scene with optional sound.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Continue an existing video with FLUX 3, using a text prompt to guide the next part of the scene with optional sound.

## Models

* **blackforestlabs/flux-3/extend-video** - Continue an existing video with a prompt describing the next part of the scene.

## Features

* Continue an existing video with a prompt describing the next part of the scene.
* Choose from `720p`, `1080p` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **5-20 seconds** - Set `duration` as an integer. Default: **5 seconds**.
* `duration` specifies the newly generated extension, not the length of the source video.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Source Video

* `video_url`: Required. Provide one publicly accessible source video URL.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `21:9`, `2:1`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `auto`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# FLUX 3 First Last Frame to Video
Source: https://docs.vidgo.ai/api-manual/video-series/blackforestlabs-flux-3-first-last-frame-to-video

api-manual/video-series/blackforestlabs-flux-3-first-last-frame-to-video.json POST /api/generate/submit
Generate a video transition between a chosen opening and closing image with FLUX 3, with text direction and optional sound.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate a video transition between a chosen opening and closing image with FLUX 3, with text direction and optional sound.

## Models

* **blackforestlabs/flux-3/first-last-frame-to-video** - Generate a transition between a starting image and an ending image.

## Features

* Generate a transition between a starting image and an ending image.
* Choose from `720p`, `1080p` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **5-20 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide exactly 2 image URLs.
* List the starting frame first and the ending frame second.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `21:9`, `2:1`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `auto`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# FLUX 3 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/blackforestlabs-flux-3-image-to-video

api-manual/video-series/blackforestlabs-flux-3-image-to-video.json POST /api/generate/submit
Bring a still image into motion with FLUX 3, using text to direct the scene with optional sound and output up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Bring a still image into motion with FLUX 3, using text to direct the scene with optional sound and output up to 1080p.

## Models

* **blackforestlabs/flux-3/image-to-video** - Generate video from image input.

## Features

* Animate a single reference image.
* Choose from `720p`, `1080p` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **5-20 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* Images must be publicly accessible and directly downloadable.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `21:9`, `2:1`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `auto`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# FLUX 3 Keyframes to Video
Source: https://docs.vidgo.ai/api-manual/video-series/blackforestlabs-flux-3-keyframes-to-video

api-manual/video-series/blackforestlabs-flux-3-keyframes-to-video.json POST /api/generate/submit
Build video sequences around chosen image keyframes with FLUX 3, controlling where visual references appear along the timeline.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Build video sequences around chosen image keyframes with FLUX 3, controlling where visual references appear along the timeline.

## Models

* **blackforestlabs/flux-3/keyframes-to-video** - Guide a video sequence with images placed at specific frame indices.

## Features

* Guide a video sequence with images placed at specific frame indices.
* Choose from `720p`, `1080p` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **5-20 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Keyframes

* `keyframes` is required and accepts **1-10 objects**.
* Each object requires a publicly accessible `image_url` and an integer `frame_index`.
* Frame indices must be unique and range from `0` to `duration * 24`, inclusive.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `21:9`, `2:1`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `auto`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# FLUX 3 Max Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/blackforestlabs-flux-3-max-image-to-video

api-manual/video-series/blackforestlabs-flux-3-max-image-to-video.json POST /api/generate/submit
The FLUX 3 Max Image to Video API animates a start image, with an optional end image, into a 5–20 second 720p or 1080p video with native audio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

FLUX 3 Max Image to Video animates a start image into a 5–20 second video at 720p or 1080p. Add a second image and the model treats it as the end frame, generating the motion that connects the two. Audio is generated in the same pass unless you turn it off.

FLUX 3 Max is a video model post-trained by fal in collaboration with Black Forest Labs on the FLUX 3 video base. Vidgo exposes it as three model IDs: [Text to Video](/api-manual/video-series/blackforestlabs-flux-3-max-text-to-video), [Image to Video](/api-manual/video-series/blackforestlabs-flux-3-max-image-to-video) and [Keyframes to Video](/api-manual/video-series/blackforestlabs-flux-3-max-keyframes-to-video).

## Models

* **blackforestlabs/flux-3-max/image-to-video** — Image-to-video with FLUX 3 Max from a start frame and an optional end frame.

## Features

* Image-to-video from one start frame.
* First and last frame control: a second image sets where the clip ends.
* Integer durations from 5 to 20 seconds at 720p or 1080p.
* Eight aspect ratios; `auto` follows the first image.
* Native audio on by default.

## Input modes

* **Start frame**: `image_urls` with one image. The video opens on that image.
* **Start and end frame**: `image_urls` with two images. The first opens the video and the second closes it.

## Required parameters

* `image_urls` (string\[]): One or two public HTTP(S) image URLs (PNG, JPEG or WebP). Order matters: start frame first, end frame second.
* `prompt` (string): How the scene moves, the camera movement and, with `sound` on, the audio. Must not be blank.

## Optional parameters

* `duration` (integer): Output length in seconds, `5`–`20`. Default `5`.
* `resolution` (string): `720p` or `1080p`. Default `720p`.
* `aspect_ratio` (string): `auto`, `21:9`, `2:1`, `16:9`, `4:3`, `1:1`, `3:4` or `9:16`. Default `auto`.
* `sound` (boolean): Generate native audio with the video. Default `true`; set `false` for silent footage.

## Output

The finished task returns one MP4 video in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Start frame only**: one portrait image with a short motion prompt, `aspect_ratio: auto` so the video keeps the image's shape.
* **Start and end frame**: a closed box and an open box as the two frames, an 8-second 1080p silent clip and a `callback_url`.

## Notes

* `image_urls` takes at most two images. To place more images at chosen points in time, use [FLUX 3 Max Keyframes to Video](/api-manual/video-series/blackforestlabs-flux-3-max-keyframes-to-video).
* Images must be publicly reachable over HTTP(S) and must not contain embedded credentials. Fields not listed above, such as `keyframes`, return 400.
* Credits are reserved at submission and depend on `resolution` and `duration`. The rate is the same for text-to-video, image-to-video and keyframes-to-video, and `sound` does not change it. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* There is no 4K tier, no `seed` and no output longer than 20 seconds on this model.


# FLUX 3 Max Keyframes to Video
Source: https://docs.vidgo.ai/api-manual/video-series/blackforestlabs-flux-3-max-keyframes-to-video

api-manual/video-series/blackforestlabs-flux-3-max-keyframes-to-video.json POST /api/generate/submit
The FLUX 3 Max Keyframes to Video API places 1–10 images on a 24 fps timeline with frame_index and generates the 5–20 second motion between them.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

FLUX 3 Max Keyframes to Video lets you pin up to 10 images to exact positions on a 24 fps timeline and generates the motion between them. It is built for storyboard animatics, product transformations and time-lapse style sequences where specific shots must appear at specific moments.

FLUX 3 Max is a video model post-trained by fal in collaboration with Black Forest Labs on the FLUX 3 video base. Vidgo exposes it as three model IDs: [Text to Video](/api-manual/video-series/blackforestlabs-flux-3-max-text-to-video), [Image to Video](/api-manual/video-series/blackforestlabs-flux-3-max-image-to-video) and [Keyframes to Video](/api-manual/video-series/blackforestlabs-flux-3-max-keyframes-to-video).

## Models

* **blackforestlabs/flux-3-max/keyframes-to-video** — Keyframes-to-video with FLUX 3 Max, 1–10 images placed by `frame_index`.

## Features

* Keyframe control with 1–10 images, each placed on a frame of a 24 fps timeline.
* Integer durations from 5 to 20 seconds at 720p or 1080p.
* Eight aspect ratios; `auto` follows the first keyframe image.
* Native audio on by default.

## Required parameters

* `keyframes` (object\[]): 1–10 objects, each with exactly two fields:
  * `image_url` (string): Public HTTP(S) URL of the keyframe image (PNG, JPEG or WebP).
  * `frame_index` (integer): Position of that image on the 24 fps timeline, from `0` to `duration × 24`. Indexes must be unique.
* `prompt` (string): The motion and transitions between keyframes, the camera movement and, with `sound` on, the audio. Must not be blank.

## Optional parameters

* `duration` (integer): Output length in seconds, `5`–`20`. Default `5`.
* `resolution` (string): `720p` or `1080p`. Default `720p`.
* `aspect_ratio` (string): `auto`, `21:9`, `2:1`, `16:9`, `4:3`, `1:1`, `3:4` or `9:16`. Default `auto`.
* `sound` (boolean): Generate native audio with the video. Default `true`; set `false` for silent footage.

## Duration

`frame_index` counts frames at 24 fps, so the last valid index depends on `duration`:

| `duration` | Valid `frame_index` range |
| - | - |
| 5 | 0–120 |
| 10 | 0–240 |
| 15 | 0–360 |
| 20 | 0–480 |

To place an image at second *t*, use `frame_index = t × 24` (for example 2.5 s → 60).

## Output

The finished task returns one MP4 video in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Three keyframes over 5 seconds**: seed, sapling and blossom at frames `0`, `60` and `120` (start, middle and end of a 5-second clip).
* **10-second 1080p storyboard with callback**: four shots at frames `0`, `96`, `168` and `240`, `21:9`, silent, with a `callback_url`.

## Notes

* A `frame_index` above `duration × 24`, a repeated index, or extra fields inside a keyframe object return 400. If you shorten `duration`, check that every index still fits.
* Fields not listed above, such as `image_urls`, return 400. For a simple start/end frame clip, use [FLUX 3 Max Image to Video](/api-manual/video-series/blackforestlabs-flux-3-max-image-to-video).
* Credits are reserved at submission and depend on `resolution` and `duration`. The rate is the same for text-to-video, image-to-video and keyframes-to-video, and `sound` does not change it. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* There is no 4K tier, no `seed` and no output longer than 20 seconds on this model.


# FLUX 3 Max Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/blackforestlabs-flux-3-max-text-to-video

api-manual/video-series/blackforestlabs-flux-3-max-text-to-video.json POST /api/generate/submit
The FLUX 3 Max Text to Video API generates 5–20 second videos from a text prompt at 720p or 1080p, with eight aspect ratios and optional native audio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

FLUX 3 Max Text to Video turns a written prompt into a 5–20 second video at 720p or 1080p, with native audio generated in the same pass. Use it for social shorts, product teasers, concept shots and storyboard drafts when you have no source image.

FLUX 3 Max is a video model post-trained by fal in collaboration with Black Forest Labs on the FLUX 3 video base. Vidgo exposes it as three model IDs: [Text to Video](/api-manual/video-series/blackforestlabs-flux-3-max-text-to-video), [Image to Video](/api-manual/video-series/blackforestlabs-flux-3-max-image-to-video) and [Keyframes to Video](/api-manual/video-series/blackforestlabs-flux-3-max-keyframes-to-video).

## Models

* **blackforestlabs/flux-3-max/text-to-video** — Text-to-video with FLUX 3 Max, 5–20 seconds, 720p or 1080p, optional audio.

## Features

* Text-to-video from a single prompt, no image needed.
* Integer durations from 5 to 20 seconds.
* 720p or 1080p output.
* Eight aspect ratios, including cinematic `21:9` and `2:1`, vertical `9:16` and square `1:1`.
* Native audio on by default; describe ambience, effects or dialogue in the prompt.

## Required parameters

* `prompt` (string): The scene, the motion and the camera movement. When `sound` is on, you can also describe the sounds you want. Must not be blank.

## Optional parameters

* `duration` (integer): Output length in seconds, `5`–`20`. Default `5`.
* `resolution` (string): `720p` or `1080p`. Default `720p`.
* `aspect_ratio` (string): `auto`, `21:9`, `2:1`, `16:9`, `4:3`, `1:1`, `3:4` or `9:16`. Default `auto`.
* `sound` (boolean): Generate native audio with the video. Default `true`; set `false` for silent footage.

## Output

The finished task returns one MP4 video in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **5-second 720p clip with audio**: a short prompt that also names the sounds (rain, traffic), `16:9`, `sound: true`.
* **15-second vertical 1080p clip, silent**: `9:16` for social feeds with `sound: false`, plus a `callback_url` to be notified on completion.

## Notes

* This ID accepts text only. To animate a start image, use [FLUX 3 Max Image to Video](/api-manual/video-series/blackforestlabs-flux-3-max-image-to-video). To pin images to specific frames, use [FLUX 3 Max Keyframes to Video](/api-manual/video-series/blackforestlabs-flux-3-max-keyframes-to-video).
* Fields not listed above, such as `image_urls` or `keyframes`, return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and `duration`. The rate is the same for text-to-video, image-to-video and keyframes-to-video, and `sound` does not change it. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* There is no 4K tier, no `seed` and no output longer than 20 seconds on this model.


# FLUX 3 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/blackforestlabs-flux-3-text-to-video

api-manual/video-series/blackforestlabs-flux-3-text-to-video.json POST /api/generate/submit
Create video scenes from text with FLUX 3, with optional sound and output up to twenty seconds at 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create video scenes from text with FLUX 3, with optional sound and output up to twenty seconds at 1080p.

## Models

* **blackforestlabs/flux-3/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **5-20 seconds** - Set `duration` as an integer. Default: **5 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `21:9`, `2:1`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`. Default: `auto`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Veo 3.1 Fast Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-veo3-1-fast-image-to-video

api-manual/video-series/google-veo3-1-fast-image-to-video.json POST /api/generate/submit
Generate motion from an image or connect opening and closing frames with Veo 3.1 Fast, with optional sound and output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate motion from an image or connect opening and closing frames with Veo 3.1 Fast, with optional sound and output up to 4K.

## Models

* **google/veo3.1-fast/image-to-video** - Generate video from image input.

## Features

* Generate motion from one starting image, or connect two images as the first and last frames.
* Choose from `720p`, `1080p`, `4k` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **4, 6, 8 seconds** - Set `duration` to one of these values. Default: **8 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide 1-2 image URLs.
* One image is used as the starting frame.
* With two images, list the starting frame first and the ending frame second.
* Images must be publicly accessible and directly downloadable.

### Generation Type

* Omit `generation_type` when supplying one image.
* With two images, `generation_type` defaults to `frame`.
* If you explicitly set `generation_type: "frame"`, exactly two images are required.

### Resolution

* `resolution`: `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `16:9`, `9:16`. Default: `16:9`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Veo 3.1 Fast Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-veo3-1-fast-text-to-video

api-manual/video-series/google-veo3-1-fast-text-to-video.json POST /api/generate/submit
Create video scenes from text with Veo 3.1 Fast, with optional sound and landscape or portrait output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create video scenes from text with Veo 3.1 Fast, with optional sound and landscape or portrait output up to 4K.

## Models

* **google/veo3.1-fast/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p`, `4k` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **4, 6, 8 seconds** - Set `duration` to one of these values. Default: **8 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Resolution

* `resolution`: `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `16:9`, `9:16`. Default: `16:9`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Veo 3.1 Official Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-veo3-1-image-to-video

api-manual/video-series/google-veo3-1-image-to-video.json POST /api/generate/submit
Generate motion from an image or connect opening and closing frames with Veo 3.1, with optional sound and output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate motion from an image or connect opening and closing frames with Veo 3.1, with optional sound and output up to 4K.

## Models

* **google/veo3.1/image-to-video** - Generate video from image input.

## Features

* Generate motion from one starting image, or connect two images as the first and last frames.
* Choose from `720p`, `1080p`, `4k` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **4, 6, 8 seconds** - Set `duration` to one of these values. Default: **8 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide 1-2 image URLs.
* One image is used as the starting frame.
* With two images, list the starting frame first and the ending frame second.
* Images must be publicly accessible and directly downloadable.

### Generation Type

* Omit `generation_type` when supplying one image.
* With two images, `generation_type` defaults to `frame`.
* If you explicitly set `generation_type: "frame"`, exactly two images are required.

### Resolution

* `resolution`: `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `16:9`, `9:16`. Default: `16:9`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Veo 3.1 Lite Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-veo3-1-lite-image-to-video

api-manual/video-series/google-veo3-1-lite-image-to-video.json POST /api/generate/submit
Generate motion from an image or connect opening and closing frames with Veo 3.1 Lite, with optional sound and output up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate motion from an image or connect opening and closing frames with Veo 3.1 Lite, with optional sound and output up to 1080p.

## Models

* **google/veo3.1-lite/image-to-video** - Generate video from image input.

## Features

* Generate motion from one starting image, or connect two images as the first and last frames.
* Choose from `720p`, `1080p` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **4, 6, 8 seconds** - Set `duration` to one of these values. Default: **8 seconds**.
* Set `duration: 8` when using `1080p` output or two input images.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide 1-2 image URLs.
* One image is used as the starting frame.
* With two images, list the starting frame first and the ending frame second.
* Images must be publicly accessible and directly downloadable.

### Generation Type

* Omit `generation_type` when supplying one image.
* With two images, `generation_type` defaults to `frame`.
* If you explicitly set `generation_type: "frame"`, exactly two images are required.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `720p`.
* `1080p` requires `duration: 8`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `16:9`, `9:16`. Default: `16:9`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Veo 3.1 Lite Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-veo3-1-lite-text-to-video

api-manual/video-series/google-veo3-1-lite-text-to-video.json POST /api/generate/submit
Create video scenes from text with Veo 3.1 Lite, with optional sound and landscape or portrait output up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create video scenes from text with Veo 3.1 Lite, with optional sound and landscape or portrait output up to 1080p.

## Models

* **google/veo3.1-lite/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **4, 6, 8 seconds** - Set `duration` to one of these values. Default: **8 seconds**.
* Set `duration: 8` when using `1080p` output.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Resolution

* `resolution`: `720p`, `1080p`. Default: `720p`.
* `1080p` requires `duration: 8`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `16:9`, `9:16`. Default: `16:9`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Veo 3.1 Official Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/google-veo3-1-text-to-video

api-manual/video-series/google-veo3-1-text-to-video.json POST /api/generate/submit
Create video scenes from text with Veo 3.1, with optional sound and landscape or portrait output up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create video scenes from text with Veo 3.1, with optional sound and landscape or portrait output up to 4K.

## Models

* **google/veo3.1/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p`, `4k` output resolutions.
* Control audio generation with `sound`; it defaults to `true`.

## Duration Options

* **4, 6, 8 seconds** - Set `duration` to one of these values. Default: **8 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion.
* The prompt must not be blank when supplied.

### Resolution

* `resolution`: `720p`, `1080p`, `4k`. Default: `720p`.

### Aspect Ratio

* `aspect_ratio`: `auto`, `16:9`, `9:16`. Default: `16:9`.

### Audio

* `sound`: Set to `true` to enable generated audio, or `false` to disable it. Default: `true`.

## Pricing

Pricing depends on the selected model and applicable output settings. Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Hailuo 02
Source: https://docs.vidgo.ai/api-manual/video-series/hailuo-02

api-manual/video-series/hailuo-02.json POST /api/generate/submit
Advanced video generation with Hailuo 02

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate high-quality videos using Hailuo 02 models. Create professional content with text-to-video and image-to-video capabilities.

## Models

* **hailuo-02** - Standard video generation (768P/512P, up to 10 seconds)
* **hailuo-02-pro** - Pro tier video generation. `resolution` accepts `512P` or `768P` (default `768P`); `duration` is not supported. The dedicated `minimax/hailuo-02/pro/*` workflow IDs below behave the same way

## Features

* **Flexible Modes**: Text-to-video and image-to-video
* **Multiple Resolutions**: 512P and 768P (Standard; the Pro workflow IDs also accept 512P and 768P)
* **Duration Options**: 6 or 10 seconds (Standard only)
* **Prompt Optimizer**: Optional AI-powered prompt enhancement
* **End Frame Control**: Specify ending frame for image-to-video (optional)

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [minimax/hailuo-02/standard/text-to-video](/api-manual/video-series/minimax-hailuo-02-standard-text-to-video)
* [minimax/hailuo-02/standard/image-to-video](/api-manual/video-series/minimax-hailuo-02-standard-image-to-video)
* [minimax/hailuo-02/pro/text-to-video](/api-manual/video-series/minimax-hailuo-02-pro-text-to-video)
* [minimax/hailuo-02/pro/image-to-video](/api-manual/video-series/minimax-hailuo-02-pro-image-to-video)


# Hailuo 2.3
Source: https://docs.vidgo.ai/api-manual/video-series/hailuo-2-3

api-manual/video-series/hailuo-2-3.json POST /api/generate/submit
Text-to-video and optional first-frame guided generation

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

`hailuo-2.3` supports prompt-only generation and optional first-frame guided generation.

## Models

* **hailuo-2.3** - Text-to-video with optional `start_image_url`

## Required parameters

* **prompt**: Text prompt for video generation

## Optional parameters

* **duration**: `6` or `10`. Default is `6`
* **resolution**: `768p` or `1080p`. Default is `768p`
* **start\_image\_url**: Optional first-frame image
* **prompt\_optimizer**: Optional boolean switch

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [minimax/hailuo-2.3/standard/text-to-video](/api-manual/video-series/minimax-hailuo-2-3-standard-text-to-video)
* [minimax/hailuo-2.3/standard/image-to-video](/api-manual/video-series/minimax-hailuo-2-3-standard-image-to-video)
* [minimax/hailuo-2.3/pro/text-to-video](/api-manual/video-series/minimax-hailuo-2-3-pro-text-to-video)
* [minimax/hailuo-2.3/pro/image-to-video](/api-manual/video-series/minimax-hailuo-2-3-pro-image-to-video)

## Notes

* `1080p` only supports `duration=6`


# Happy Horse
Source: https://docs.vidgo.ai/api-manual/video-series/happy-horse

api-manual/video-series/happy-horse.json POST /api/generate/submit
Alibaba Happy Horse text-to-video, image-to-video, reference-to-video, and video-edit workflows

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

`happy-horse` supports text-to-video, image-to-video, reference-to-video, and video-edit through one public model ID.

## Models

* **happy-horse** - Alibaba Happy Horse 1.0 video generation and editing

## Required parameters

* **model**: `happy-horse`
* **input.prompt**: Required for text-to-video, reference-to-video and video-edit (maximum 2,500 characters). Optional for image-to-video.
* **input.video\_url**: Required for video-edit.
* **input.reference\_image\_urls**: Required for reference-to-video.

## Optional parameters

* **image\_urls**: Single-item first-frame image URL array for image-to-video
* **reference\_image\_urls**: 1-9 images for reference-to-video, or up to 5 optional images for video-edit
* **aspect\_ratio**: Text-to-video and reference-to-video only. `16:9`, `9:16`, `1:1`, `4:3`, or `3:4`. Default is `16:9`
* **resolution**: `720p` or `1080p`. Default is `1080p`
* **duration**: Integer seconds from `3` to `15` for text-to-video, image-to-video, and reference-to-video. Ignored for video-edit
* **audio\_setting**: Video-edit only. `auto` or `origin`. Default is `auto`
* **seed**: Optional integer from `0` to `2147483647`
* **enable\_safety\_checker**: Optional boolean

## Workflows

### Text to Video

Send a `prompt` without image or video input fields. Use `aspect_ratio`, `resolution`, and `duration` to control the output.

### Image to Video

Send a single-item `image_urls` array. The image is used as the first frame. `prompt` is optional; when supplied it must be nonblank and at most 2,500 characters.

### Reference to Video

Send `reference_image_urls` with 1-9 images and a required `prompt`. Describe the referenced subjects in the prompt; `character1` and `character2` markers are not required.

Do not combine `reference_image_urls` with `image_urls`.

### Video Edit

Send `video_url` with a required edit `prompt`. Optional `reference_image_urls` (up to 5) can guide the edit; describe the referenced subjects in the prompt, no special markers are required. `audio_setting` controls whether Happy Horse decides audio handling automatically or preserves the original audio.

The source video must be 3-60 seconds. Video-edit billing uses the probed source video duration according to the current workflow pricing configuration.

## Pricing

Happy Horse billing depends on the selected workflow, output resolution, and billable video duration. Check the Vidgo pricing page or dashboard for the current rate before submitting production traffic.

Text-to-video, image-to-video, and reference-to-video use the requested `duration`. Video-edit uses the probed source video duration, capped at the maximum billable duration for this workflow.

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [alibaba/happyhorse-1.0/text-to-video](/api-manual/video-series/alibaba-happyhorse-1-0-text-to-video)
* [alibaba/happyhorse-1.0/image-to-video](/api-manual/video-series/alibaba-happyhorse-1-0-image-to-video)
* [alibaba/happyhorse-1.0/reference-to-video](/api-manual/video-series/alibaba-happyhorse-1-0-reference-to-video)
* [alibaba/happyhorse-1.0/video-edit](/api-manual/video-series/alibaba-happyhorse-1-0-video-edit)

## Notes

* `image_urls` must resolve to exactly one image and is only for image-to-video.
* `reference_image_urls` selects reference-to-video unless `video_url` is present, in which case it is treated as optional video-edit reference imagery.
* For text-to-video and reference-to-video, omit `video_url` and use `aspect_ratio`.
* For video-edit, omit `duration`; the source video duration is used for validation and duration-based billing.


# Kling 1.6
Source: https://docs.vidgo.ai/api-manual/video-series/kling-1-6

api-manual/video-series/kling-1-6.json POST /api/generate/submit
Kling 1.6 Standard and Pro video generation through the Vidgo API

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Input Routing

The supplied image fields determine the generation workflow.
All Kling 1.6 requests require `prompt` and `duration`.

* Provide no image fields to generate text-to-video.
* Provide `start_image_url` to generate image-to-video.
* Provide `start_image_url` and `end_image_url` with `kling-1.6/pro` to control both the first and last frame.
* Provide `image_urls` with 1 to 4 reference images to generate Elements.

`image_urls` cannot be used together with `start_image_url`, `end_image_url`, or `cfg_scale`.

## Parameter Compatibility

| Model and workflow | How Vidgo detects it | Supported parameters |
| - | - | - |
| Text to Video (`standard` or `pro`) | No image fields | `prompt`, `duration`, `aspect_ratio`, `negative_prompt`, `cfg_scale` |
| Standard Image to Video | `kling-1.6/standard` with `start_image_url` | `prompt`, `duration`, `start_image_url`, `negative_prompt`, `cfg_scale` |
| Pro Image to Video | `kling-1.6/pro` with `start_image_url` and optional `end_image_url` | `prompt`, `duration`, `aspect_ratio`, `start_image_url`, `end_image_url`, `negative_prompt`, `cfg_scale` |
| Elements (`standard` or `pro`) | `image_urls` | `prompt`, `duration`, `aspect_ratio`, `negative_prompt`, `image_urls` |

Notes:

* `end_image_url` requires `start_image_url` and is only supported by `kling-1.6/pro`.
* `kling-1.6/standard` image-to-video does not support `aspect_ratio`.
* Elements supports 1 to 4 `image_urls` and does not support `cfg_scale`.
* `prompt` and `negative_prompt` support up to 2500 characters.
* `cfg_scale` must be between 0 and 1 when provided.

## Response and Status

The submit endpoint returns a Vidgo `task_id`. Use it with the unified Query Task Status endpoint to retrieve the final video URL.

Initial submit responses include `status`, which is usually `running` after the upstream video task has been created. If submission fails during upstream task creation, the status may be `failed` and the task can be inspected through the status endpoint.

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [kwaivgi/kling-video/v1.6/standard/text-to-video](/api-manual/video-series/kwaivgi-kling-video-v1-6-standard-text-to-video)
* [kwaivgi/kling-video/v1.6/standard/image-to-video](/api-manual/video-series/kwaivgi-kling-video-v1-6-standard-image-to-video)
* [kwaivgi/kling-video/v1.6/pro/text-to-video](/api-manual/video-series/kwaivgi-kling-video-v1-6-pro-text-to-video)
* [kwaivgi/kling-video/v1.6/pro/image-to-video](/api-manual/video-series/kwaivgi-kling-video-v1-6-pro-image-to-video)


# Kling 2.1
Source: https://docs.vidgo.ai/api-manual/video-series/kling-2-1

api-manual/video-series/kling-2-1.json POST /api/generate/submit
Standard and Pro image-to-video generation

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Kling 2.1 provides two public models under the same request format: `kling-2.1/standard` and `kling-2.1/pro`.

## Models

* **kling-2.1/standard** - Start-frame guided image-to-video
* **kling-2.1/pro** - Start-frame guided image-to-video with optional end frame

## Required parameters

* **prompt**: Text prompt for generation
* **start\_image\_url**: Required first frame image

## Optional parameters

* **duration**: `5` or `10`. Default is `5`
* **end\_image\_url**: Optional, only supported by `kling-2.1/pro`
* **negative\_prompt**: Optional negative prompt

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [kwaivgi/kling-video/v2.1/standard/image-to-video](/api-manual/video-series/kwaivgi-kling-video-v2-1-standard-image-to-video)
* [kwaivgi/kling-video/v2.1/pro/image-to-video](/api-manual/video-series/kwaivgi-kling-video-v2-1-pro-image-to-video)

## Notes

* `kling-2.1/standard` does not support `end_image_url`
* `kling-2.1/pro` supports both start image and optional end image


# Kling 2.6 Motion Control
Source: https://docs.vidgo.ai/api-manual/video-series/kling-2.6-motion-control

api-manual/video-series/kling-2.6-motion-control.json POST /api/generate/submit
Advanced motion transfer with character animation

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Transfer motion from reference videos to character images with precise motion control. Create dynamic character animations by combining static images with motion reference videos while controlling scene details through text prompts.

## Models

* **kling-2.6-motion-control** - Motion transfer from video to character image

## Features

* **Motion Transfer**: Transfer real human motion, gestures, and expressions from reference videos to character images
* **Character Animation**: Animate static characters with dynamic motion patterns
* **Character Orientation Control**: Choose between matching image orientation (max 10s output) or video orientation (max 30s output)
* **Scene Control**: Control scene details and environment through optional text prompts
* **Flexible Resolution**: Choose between 720p (standard) or 1080p (high quality) output

## Required parameters

* **image\_urls**: Single character image showing the subject's head, shoulders, and torso
* **video\_urls**: Single reference video (3-30 seconds) for motion transfer
* **character\_orientation**:
  * `"image"` - Matches the person's orientation in the photograph (maximum 10-second output)
  * `"video"` - Maintains consistency with character orientation from the reference video (maximum 30-second output)
* **mode**: Output resolution (`"720p"` or `"1080p"`)

## Optional parameters

* **prompt**: Text description of the desired output scene (max 2,500 characters)

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [kwaivgi/kling-v2.6-std/motion-control](/api-manual/video-series/kwaivgi-kling-v2-6-std-motion-control)
* [kwaivgi/kling-v2.6-pro/motion-control](/api-manual/video-series/kwaivgi-kling-v2-6-pro-motion-control)


# Kling 3.0
Source: https://docs.vidgo.ai/api-manual/video-series/kling-3-0

api-manual/video-series/kling-3-0.json POST /api/generate/submit
Next-generation video generation with standard and pro resolution

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **kling-3.0/standard** - HD resolution (\~720p/1K)
* **kling-3.0/pro** - Full HD resolution (1080p/2K), sharper and more detailed output

## Features

* **Native Audio**: Controlled by `input.sound`. Set `true` to enable sound effects and `false` to disable them. When `multi_shots` is `true`, `sound` must be `true`
* **Flexible Duration**: 3 to 15 seconds
* **Aspect Ratios**: 1:1 (square), 16:9 (landscape), and 9:16 (portrait)
* **Modes**: Text-to-video and image-to-video (start and end frames)
* **Multi-Shot**: Multi-shot storytelling with per-shot prompts and durations
* **Element References**: Define reusable elements (via images or video) and reference them in prompts using `@element_name` syntax. `image_urls` is required when using element references

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [kwaivgi/kling-v3.0-std/text-to-video](/api-manual/video-series/kwaivgi-kling-v3-0-std-text-to-video)
* [kwaivgi/kling-v3.0-std/image-to-video](/api-manual/video-series/kwaivgi-kling-v3-0-std-image-to-video)
* [kwaivgi/kling-v3.0-pro/text-to-video](/api-manual/video-series/kwaivgi-kling-v3-0-pro-text-to-video)
* [kwaivgi/kling-v3.0-pro/image-to-video](/api-manual/video-series/kwaivgi-kling-v3-0-pro-image-to-video)


# Kling 3.0 Motion Control
Source: https://docs.vidgo.ai/api-manual/video-series/kling-3-0-motion-control

api-manual/video-series/kling-3-0-motion-control.json POST /api/generate/submit
Reference image and reference video motion transfer

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

`kling-3.0-motion-control` combines one reference image and one reference video to transfer motion to the target character.

## Models

* **kling-3.0-motion-control** - Motion transfer from reference video to reference image

## Required parameters

* **image\_urls**: Exactly `1` reference image
* **video\_urls**: Exactly `1` reference video
* **character\_orientation**: `image` or `video`

## Optional parameters

* **prompt**: Optional scene prompt
* **resolution**: `720p` or `1080p`. Default is `720p`

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [kwaivgi/kling-v3.0-std/motion-control](/api-manual/video-series/kwaivgi-kling-v3-0-std-motion-control)
* [kwaivgi/kling-v3.0-pro/motion-control](/api-manual/video-series/kwaivgi-kling-v3-0-pro-motion-control)

## Notes

* Images support `.jpg`, `.jpeg`, `.png`, up to `10MB`
* Videos support `.mp4`, `.mov`, up to `100MB`
* Reference video must be at least `3` seconds
* When `character_orientation=image`, the reference video must not exceed `10` seconds
* When `character_orientation=video`, the reference video must not exceed `30` seconds


# Kling 2.5 Turbo Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v2-5-turbo-pro-image-to-video

api-manual/video-series/kwaivgi-kling-v2-5-turbo-pro-image-to-video.json POST /api/generate/submit
Generate video with Kling 2.5 Turbo Pro using a start frame, an end frame, or both, plus a text prompt.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate video with Kling 2.5 Turbo Pro using a start frame, an end frame, or both, plus a text prompt.

## Models

* **kwaivgi/kling-v2.5-turbo-pro/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Guide generation with `start_image_url`, `end_image_url`, or both.

## Duration Options

* **5, 10 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**. Strings, booleans, fractional durations, and `null` are rejected.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials or whitespace. Image fields do not accept arrays or Base64.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-2,500 characters**.
* Supply a string, not a number, boolean, array, or object. Whitespace is trimmed before the 2,500 Unicode character limit is checked.

### Start Image URL

* `start_image_url`: First-frame public HTTP(S) URL. Required only when `end_image_url` is omitted.

### End Image URL

* `end_image_url`: Last-frame public HTTP(S) URL. Required only when `start_image_url` is omitted.
* Provide at least one frame: start only, end only, or both. Omit unused URL fields; empty strings and `null` are rejected.

### Aspect Ratio

* `aspect_ratio`: Optional string. No preset choices or default value are specified. Omit when unused.

### Negative Prompt

* `negative_prompt`: Optional. Describe content to avoid.

## Billing Factors

Billing depends on video duration (`duration`). Text-to-video and image-to-video use the same rates.

Credits are deducted on submission and refunded if the task fails.


# Kling 2.5 Turbo Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v2-5-turbo-pro-text-to-video

api-manual/video-series/kwaivgi-kling-v2-5-turbo-pro-text-to-video.json POST /api/generate/submit
Create five- or ten-second scenes with Kling 2.5 Turbo Pro, using text to direct subjects, action, and visual style.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create five- or ten-second scenes with Kling 2.5 Turbo Pro, using text to direct subjects, action, and visual style.

## Models

* **kwaivgi/kling-v2.5-turbo-pro/text-to-video** - Generate video scenes from text prompts.
* For frame guidance, use the [image-to-video endpoint](/api-manual/video-series/kwaivgi-kling-v2-5-turbo-pro-image-to-video).

## Features

* Generate video scenes from text prompts.
* Use `negative_prompt` to describe content to avoid.

## Duration Options

* **5, 10 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds**. Strings, booleans, fractional durations, and `null` are rejected.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-2,500 characters**.
* Supply a string, not a number, boolean, array, or object. Whitespace is trimmed before the 2,500 Unicode character limit is checked.

### Aspect Ratio

* `aspect_ratio`: Optional string. No preset choices or default value are specified. Omit when unused.

### Negative Prompt

* `negative_prompt`: Optional. Describe content to avoid.

## Billing Factors

Billing depends on video duration (`duration`). Text-to-video and image-to-video use the same rates.

Credits are deducted on submission and refunded if the task fails.


# Kling 2.6 Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v2-6-pro-image-to-video

api-manual/video-series/kwaivgi-kling-v2-6-pro-image-to-video.json POST /api/generate/submit
Kling 2.6 Pro Image to Video API parameters and billing factors

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Submit `kwaivgi/kling-v2.6-pro/image-to-video` through `POST /api/generate/submit`. Query the returned `task_id` with [Task Status](/api-manual/task-management/status), or provide a top-level `callback_url`.

## Parameters

* `prompt`: Required nonblank string, 1 to 1,000 characters after trimming leading and trailing whitespace.
* `duration`: Required integer: 5 or 10 seconds. Strings and fractional durations are rejected.
* `aspect_ratio`: Required: 16:9, 9:16 or 1:1. This field must be submitted but does not control the output aspect ratio for image-to-video.
* `sound`: Required boolean: true for audio, false for no audio. Must be false when end\_image\_url is provided.
* `image_urls`: Required array containing at least one publicly accessible HTTP(S) image URL. URLs must not contain embedded credentials or whitespace.
* `end_image_url`: Optional publicly accessible HTTP(S) last-frame image URL, without embedded credentials or whitespace. Omit this field when not using a last frame; empty strings and null are not accepted. Requires sound=false.
* Use standard JSON types. Unknown input fields and invalid values are rejected before submission and charging.
* URLs using `example.com` or `your-domain.com` are placeholders. Replace them with your own publicly accessible image URLs and callback endpoint where applicable.

## Billing Factors

Billing depends on video duration (`duration`) and whether audio is generated (`sound`). The Pro text-to-video and image-to-video endpoints use the same rates.


# Kling 2.6 Pro Motion Control
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v2-6-pro-motion-control

api-manual/video-series/kwaivgi-kling-v2-6-pro-motion-control.json POST /api/generate/submit
Transfer movement from a reference video to an image subject with Kling 2.6 Pro, with orientation control and 1080p output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Transfer movement from a reference video to an image subject with Kling 2.6 Pro, with orientation control and 1080p output.

## Models

* **kwaivgi/kling-v2.6-pro/motion-control** - Transfer movement from a reference video to a character in a reference image.

## Features

* Transfer movement from a reference video to a character in a reference image.
* Output resolution is fixed to `1080p` for this model ID.

## Duration Options

* **Image orientation** - The reference video must be **3-10 seconds** when `character_orientation` is `image`.
* **Video orientation** - The reference video must be **3-30 seconds** when `character_orientation` is `video`.
* The service inspects the reference video duration.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt` is optional. `null`, empty, and whitespace-only values are treated as omitted.
* Nonempty prompts are trimmed and limited to **2,500 characters**.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* The image provides the target character for motion transfer.
* Images must be publicly accessible and directly downloadable.

### Source Video

* `video_urls`: Required. Provide an array containing exactly one publicly accessible source video URL.
* The video provides the movement to transfer. Its duration must match the orientation limits above.
* The source video must be directly downloadable over HTTP(S) and no larger than **100 MiB** for duration inspection.

### Character Orientation

* `character_orientation` is required: choose `image` or `video`.
* `image` allows reference videos from **3 to 10 seconds**.
* `video` allows reference videos from **3 to 30 seconds**.

### Resolution

* `resolution`: `1080p`. Default: `1080p`.
* This model ID fixes the resolution. Omit the field to use its default; a resolution from the other tier is rejected.

## Pricing

Credits are calculated from the detected reference-video duration multiplied by this endpoint's rate, then rounded up to a whole credit.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 2.6 Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v2-6-pro-text-to-video

api-manual/video-series/kwaivgi-kling-v2-6-pro-text-to-video.json POST /api/generate/submit
Kling 2.6 Pro Text to Video API parameters and billing factors

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Submit `kwaivgi/kling-v2.6-pro/text-to-video` through `POST /api/generate/submit`. Query the returned `task_id` with [Task Status](/api-manual/task-management/status), or provide a top-level `callback_url`.

## Parameters

* `prompt`: Required nonblank string, 1 to 1,000 characters after trimming leading and trailing whitespace.
* `duration`: Required integer: 5 or 10 seconds. Strings and fractional durations are rejected.
* `aspect_ratio`: Required: 16:9, 9:16 or 1:1. Controls the output video aspect ratio.
* `sound`: Required boolean: true for audio, false for no audio.
* Use standard JSON types. Unknown input fields and invalid values are rejected before submission and charging.
* URLs using `example.com` or `your-domain.com` are placeholders. Replace them with your own publicly accessible image URLs and callback endpoint where applicable.

## Billing Factors

Billing depends on video duration (`duration`) and whether audio is generated (`sound`). The Pro text-to-video and image-to-video endpoints use the same rates.


# Kling 2.6 Standard Motion Control
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v2-6-std-motion-control

api-manual/video-series/kwaivgi-kling-v2-6-std-motion-control.json POST /api/generate/submit
Transfer movement from a reference video to an image subject with Kling 2.6 Standard, with orientation control and 720p output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Transfer movement from a reference video to an image subject with Kling 2.6 Standard, with orientation control and 720p output.

## Models

* **kwaivgi/kling-v2.6-std/motion-control** - Transfer movement from a reference video to a character in a reference image.

## Features

* Transfer movement from a reference video to a character in a reference image.
* Output resolution is fixed to `720p` for this model ID.

## Duration Options

* **Image orientation** - The reference video must be **3-10 seconds** when `character_orientation` is `image`.
* **Video orientation** - The reference video must be **3-30 seconds** when `character_orientation` is `video`.
* The service inspects the reference video duration.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt` is optional. `null`, empty, and whitespace-only values are treated as omitted.
* Nonempty prompts are trimmed and limited to **2,500 characters**.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* The image provides the target character for motion transfer.
* Images must be publicly accessible and directly downloadable.

### Source Video

* `video_urls`: Required. Provide an array containing exactly one publicly accessible source video URL.
* The video provides the movement to transfer. Its duration must match the orientation limits above.
* The source video must be directly downloadable over HTTP(S) and no larger than **100 MiB** for duration inspection.

### Character Orientation

* `character_orientation` is required: choose `image` or `video`.
* `image` allows reference videos from **3 to 10 seconds**.
* `video` allows reference videos from **3 to 30 seconds**.

### Resolution

* `resolution`: `720p`. Default: `720p`.
* This model ID fixes the resolution. Omit the field to use its default; a resolution from the other tier is rejected.

## Pricing

Credits are calculated from the detected reference-video duration multiplied by this endpoint's rate, then rounded up to a whole credit.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 3.0 4K Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-0-4k-image-to-video

api-manual/video-series/kwaivgi-kling-v3-0-4k-image-to-video.json POST /api/generate/submit
Animate image references with Kling 3.0 4K, with support for multiple shots, individual shot prompts, and optional sound.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate image references with Kling 3.0 4K, with support for multiple shots, individual shot prompts, and optional sound.

## Models

* **kwaivgi/kling-v3.0-4k/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Provide 1–2 images in `image_urls`.
* Describe individual shots with `multi_prompt` and enable `multi_shots`.
* Supply additional subject references through `kling_elements`.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Required.
* With `multi_prompt`, the shot durations must add up to the requested `duration`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below. Send durations as JSON integers and switches as JSON booleans. Use public HTTP(S) media URLs with a hostname and URL-based access.

* Each `multi_prompt` shot requires a nonblank `prompt` and an integer `duration` from **1 to 12 seconds**.
* The shot durations must sum to the requested `duration`. Set `multi_shots: true` and `sound: true` when using `multi_prompt`.
* When supplying `kling_elements`, provide `image_urls` and reference the element by `@element_name` in the prompt.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1–2 items: the start frame first, followed by an optional end frame. Each item must be an HTTP(S) URL.

### Prompt

* `prompt`: Required when `multi_shots` is `false`. Describe the scene and desired motion. Length: **1-2,500 characters**.
* Use nonblank text for each prompt.

### Multi-Prompt

* `multi_prompt`: Required when `multi_shots` is `true`. Supply a nonempty array. Each item is an object.

* `multi_prompt[].prompt`: Required. Maximum length: **2,500 characters**.

* `multi_prompt[].duration`: Required. Duration in seconds. Integer from `1` to `12`.

* Each shot duration is measured in seconds.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`. Default: `false`.

### Sound

* `sound`: Required. Set `true` to generate audio or `false` for a silent single shot. Multiple shots require `true`.

### Kling Elements

* `kling_elements`: Optional. Provide element-reference objects. Supply an array. Each item is an object.
* Element object properties are preserved in the submitted request.


# Kling 3.0 4K Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-0-4k-text-to-video

api-manual/video-series/kwaivgi-kling-v3-0-4k-text-to-video.json POST /api/generate/submit
Direct video sequences with Kling 3.0 4K, using individual shot prompts, shot timing, and optional sound to shape a scene.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Direct video sequences with Kling 3.0 4K, using individual shot prompts, shot timing, and optional sound to shape a scene.

## Models

* **kwaivgi/kling-v3.0-4k/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Describe individual shots with `multi_prompt` and enable `multi_shots`.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Required.
* With `multi_prompt`, the shot durations must add up to the requested `duration`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below. Send durations as JSON integers and switches as JSON booleans.

* Each `multi_prompt` shot requires a nonblank `prompt` and an integer `duration` from **1 to 12 seconds**.
* The shot durations must sum to the requested `duration`. Set `multi_shots: true` and `sound: true` when using `multi_prompt`.

### Prompt

* `prompt`: Required when `multi_shots` is `false`. Describe the scene and desired motion. Length: **1-2,500 characters**.
* Use nonblank text for each prompt.

### Multi-Prompt

* `multi_prompt`: Required when `multi_shots` is `true`. Supply a nonempty array. Each item is an object.

* `multi_prompt[].prompt`: Required. Maximum length: **2,500 characters**.

* `multi_prompt[].duration`: Required. Duration in seconds. Integer from `1` to `12`.

* Each shot duration is measured in seconds.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`. Default: `false`.

### Sound

* `sound`: Required. Set `true` to generate audio or `false` for a silent single shot. Multiple shots require `true`.


# Kling 3.0 Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-0-pro-image-to-video

api-manual/video-series/kwaivgi-kling-v3-0-pro-image-to-video.json POST /api/generate/submit
Animate image references with Kling 3.0 Pro, with support for multiple shots, individual shot prompts, and optional sound.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate image references with Kling 3.0 Pro, with support for multiple shots, individual shot prompts, and optional sound.

## Models

* **kwaivgi/kling-v3.0-pro/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Provide 1–2 images in `image_urls`.
* Describe individual shots with `multi_prompt` and enable `multi_shots`.
* Supply additional subject references through `kling_elements`.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Required.
* With `multi_prompt`, the shot durations must add up to the requested `duration`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below. Send durations as JSON integers and switches as JSON booleans. Use public HTTP(S) media URLs with a hostname and URL-based access.

* Each `multi_prompt` shot requires a nonblank `prompt` and an integer `duration` from **1 to 12 seconds**.
* The shot durations must sum to the requested `duration`. Set `multi_shots: true` and `sound: true` when using `multi_prompt`.
* When supplying `kling_elements`, provide `image_urls` and reference the element by `@element_name` in the prompt.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1–2 items: the start frame first, followed by an optional end frame. For multiple shots, provide one start frame. Each item must be an HTTP(S) URL.

### Prompt

* `prompt`: Required when `multi_shots` is `false`. Describe the scene and desired motion. Length: **1-2,500 characters**.
* Use nonblank text for each prompt.

### Multi-Prompt

* `multi_prompt`: Required when `multi_shots` is `true`. Supply a nonempty array. Each item is an object.

* `multi_prompt[].prompt`: Required. Maximum length: **2,500 characters**.

* `multi_prompt[].duration`: Required. Duration in seconds. Integer from `1` to `12`.

* Each shot duration is measured in seconds.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`. Default: `false`.

### Sound

* `sound`: Optional. Set `true` to generate audio or `false` for a silent single shot. Multiple shots require `true`. Default: `true`.

### Kling Elements

* `kling_elements`: Optional. Provide element-reference objects. Supply an array. Each item is an object.

* `kling_elements[].name`: Required. Nonblank element name used in `@element_name` references.

* `kling_elements[].description`: Optional. String.

* `kling_elements[].element_input_urls`: Provide 2–4 JPG/PNG image URLs, up to 10 MB per image, when defining the element with images.

* `kling_elements[].element_input_video_urls`: Provide one MP4/MOV video URL, up to 50 MB, when defining the element with video. Choose one media type per element.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `16:9`, `9:16`. The first frame determines the output aspect ratio.


# Kling 3.0 Pro Motion Control
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-0-pro-motion-control

api-manual/video-series/kwaivgi-kling-v3-0-pro-motion-control.json POST /api/generate/submit
Transfer movement from a reference video to an image subject with Kling 3.0 Pro, with orientation control and 1080p output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Transfer movement from a reference video to an image subject with Kling 3.0 Pro, with orientation control and 1080p output.

## Models

* **kwaivgi/kling-v3.0-pro/motion-control** - Transfer movement from a reference video to a character in a reference image.

## Features

* Transfer movement from a reference video to a character in a reference image.
* Output resolution is fixed to `1080p` for this model ID.
* Add one facial element when using video orientation.

## Duration Options

* **Image orientation** - The reference video must be **3-10 seconds** when `character_orientation` is `image`.
* **Video orientation** - The reference video must be **3-30 seconds** when `character_orientation` is `video`.
* The service inspects the reference video duration.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt` is optional. `null`, empty, and whitespace-only values are treated as omitted.
* Nonempty prompts are trimmed and limited to **2,500 characters**.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* The image provides the target character for motion transfer.
* Images must be publicly accessible and directly downloadable.

### Source Video

* `video_urls`: Required. Provide an array containing exactly one publicly accessible source video URL.
* The video provides the movement to transfer. Its duration must match the orientation limits above.
* The source video must be directly downloadable over HTTP(S) and no larger than **100 MiB** for duration inspection.

### Character Orientation

* `character_orientation` is required: choose `image` or `video`.
* `image` allows reference videos from **3 to 10 seconds**.
* `video` allows reference videos from **3 to 30 seconds**.

### Facial Elements

* `kling_elements` is optional and accepts at most **one facial element**.
* Use it only when `character_orientation` is `video`.
* The element requires `frontal_image_url` and **1-3** `reference_image_urls`.

### Resolution

* `resolution`: `1080p`. Default: `1080p`.
* This model ID fixes the resolution. Omit the field to use its default; a resolution from the other tier is rejected.

## Pricing

Credits are calculated from the detected reference-video duration multiplied by this endpoint's rate, then rounded up to a whole credit.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 3.0 Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-0-pro-text-to-video

api-manual/video-series/kwaivgi-kling-v3-0-pro-text-to-video.json POST /api/generate/submit
Direct video sequences with Kling 3.0 Pro, using individual shot prompts, shot timing, and optional sound to shape a scene.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Direct video sequences with Kling 3.0 Pro, using individual shot prompts, shot timing, and optional sound to shape a scene.

## Models

* **kwaivgi/kling-v3.0-pro/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Describe individual shots with `multi_prompt` and enable `multi_shots`.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Required.
* With `multi_prompt`, the shot durations must add up to the requested `duration`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below. Send durations as JSON integers and switches as JSON booleans.

* Each `multi_prompt` shot requires a nonblank `prompt` and an integer `duration` from **1 to 12 seconds**.
* The shot durations must sum to the requested `duration`. Set `multi_shots: true` and `sound: true` when using `multi_prompt`.

### Prompt

* `prompt`: Required when `multi_shots` is `false`. Describe the scene and desired motion. Length: **1-2,500 characters**.
* Use nonblank text for each prompt.

### Multi-Prompt

* `multi_prompt`: Required when `multi_shots` is `true`. Supply a nonempty array. Each item is an object.

* `multi_prompt[].prompt`: Required. Maximum length: **2,500 characters**.

* `multi_prompt[].duration`: Required. Duration in seconds. Integer from `1` to `12`.

* Each shot duration is measured in seconds.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`. Default: `false`.

### Sound

* `sound`: Optional. Set `true` to generate audio or `false` for a silent single shot. Multiple shots require `true`. Default: `true`.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `16:9`, `9:16`. Default: `1:1`.


# Kling 3.0 Standard Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-0-std-image-to-video

api-manual/video-series/kwaivgi-kling-v3-0-std-image-to-video.json POST /api/generate/submit
Animate image references with Kling 3.0 Standard, with support for multiple shots, individual shot prompts, and optional sound.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate image references with Kling 3.0 Standard, with support for multiple shots, individual shot prompts, and optional sound.

## Models

* **kwaivgi/kling-v3.0-std/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Provide 1–2 images in `image_urls`.
* Describe individual shots with `multi_prompt` and enable `multi_shots`.
* Supply additional subject references through `kling_elements`.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Required.
* With `multi_prompt`, the shot durations must add up to the requested `duration`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below. Send durations as JSON integers and switches as JSON booleans. Use public HTTP(S) media URLs with a hostname and URL-based access.

* Each `multi_prompt` shot requires a nonblank `prompt` and an integer `duration` from **1 to 12 seconds**.
* The shot durations must sum to the requested `duration`. Set `multi_shots: true` and `sound: true` when using `multi_prompt`.
* When supplying `kling_elements`, provide `image_urls` and reference the element by `@element_name` in the prompt.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1–2 items: the start frame first, followed by an optional end frame. For multiple shots, provide one start frame. Each item must be an HTTP(S) URL.

### Prompt

* `prompt`: Required when `multi_shots` is `false`. Describe the scene and desired motion. Length: **1-2,500 characters**.
* Use nonblank text for each prompt.

### Multi-Prompt

* `multi_prompt`: Required when `multi_shots` is `true`. Supply a nonempty array. Each item is an object.

* `multi_prompt[].prompt`: Required. Maximum length: **2,500 characters**.

* `multi_prompt[].duration`: Required. Duration in seconds. Integer from `1` to `12`.

* Each shot duration is measured in seconds.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`. Default: `false`.

### Sound

* `sound`: Optional. Set `true` to generate audio or `false` for a silent single shot. Multiple shots require `true`. Default: `true`.

### Kling Elements

* `kling_elements`: Optional. Provide element-reference objects. Supply an array. Each item is an object.

* `kling_elements[].name`: Required. Nonblank element name used in `@element_name` references.

* `kling_elements[].description`: Optional. String.

* `kling_elements[].element_input_urls`: Provide 2–4 JPG/PNG image URLs, up to 10 MB per image, when defining the element with images.

* `kling_elements[].element_input_video_urls`: Provide one MP4/MOV video URL, up to 50 MB, when defining the element with video. Choose one media type per element.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `16:9`, `9:16`. The first frame determines the output aspect ratio.


# Kling 3.0 Standard Motion Control
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-0-std-motion-control

api-manual/video-series/kwaivgi-kling-v3-0-std-motion-control.json POST /api/generate/submit
Transfer movement from a reference video to an image subject with Kling 3.0 Standard, with orientation control and 720p output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Transfer movement from a reference video to an image subject with Kling 3.0 Standard, with orientation control and 720p output.

## Models

* **kwaivgi/kling-v3.0-std/motion-control** - Transfer movement from a reference video to a character in a reference image.

## Features

* Transfer movement from a reference video to a character in a reference image.
* Output resolution is fixed to `720p` for this model ID.
* Add one facial element when using video orientation.

## Duration Options

* **Image orientation** - The reference video must be **3-10 seconds** when `character_orientation` is `image`.
* **Video orientation** - The reference video must be **3-30 seconds** when `character_orientation` is `video`.
* The service inspects the reference video duration.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt` is optional. `null`, empty, and whitespace-only values are treated as omitted.
* Nonempty prompts are trimmed and limited to **2,500 characters**.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* The image provides the target character for motion transfer.
* Images must be publicly accessible and directly downloadable.

### Source Video

* `video_urls`: Required. Provide an array containing exactly one publicly accessible source video URL.
* The video provides the movement to transfer. Its duration must match the orientation limits above.
* The source video must be directly downloadable over HTTP(S) and no larger than **100 MiB** for duration inspection.

### Character Orientation

* `character_orientation` is required: choose `image` or `video`.
* `image` allows reference videos from **3 to 10 seconds**.
* `video` allows reference videos from **3 to 30 seconds**.

### Facial Elements

* `kling_elements` is optional and accepts at most **one facial element**.
* Use it only when `character_orientation` is `video`.
* The element requires `frontal_image_url` and **1-3** `reference_image_urls`.

### Resolution

* `resolution`: `720p`. Default: `720p`.
* This model ID fixes the resolution. Omit the field to use its default; a resolution from the other tier is rejected.

## Pricing

Credits are calculated from the detected reference-video duration multiplied by this endpoint's rate, then rounded up to a whole credit.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 3.0 Standard Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-0-std-text-to-video

api-manual/video-series/kwaivgi-kling-v3-0-std-text-to-video.json POST /api/generate/submit
Direct video sequences with Kling 3.0 Standard, using individual shot prompts, shot timing, and optional sound to shape a scene.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Direct video sequences with Kling 3.0 Standard, using individual shot prompts, shot timing, and optional sound to shape a scene.

## Models

* **kwaivgi/kling-v3.0-std/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Describe individual shots with `multi_prompt` and enable `multi_shots`.

## Duration Options

* **3-15 seconds** - Set `duration` as an integer. Required.
* With `multi_prompt`, the shot durations must add up to the requested `duration`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below. Send durations as JSON integers and switches as JSON booleans.

* Each `multi_prompt` shot requires a nonblank `prompt` and an integer `duration` from **1 to 12 seconds**.
* The shot durations must sum to the requested `duration`. Set `multi_shots: true` and `sound: true` when using `multi_prompt`.

### Prompt

* `prompt`: Required when `multi_shots` is `false`. Describe the scene and desired motion. Length: **1-2,500 characters**.
* Use nonblank text for each prompt.

### Multi-Prompt

* `multi_prompt`: Required when `multi_shots` is `true`. Supply a nonempty array. Each item is an object.

* `multi_prompt[].prompt`: Required. Maximum length: **2,500 characters**.

* `multi_prompt[].duration`: Required. Duration in seconds. Integer from `1` to `12`.

* Each shot duration is measured in seconds.

### Multi-Shot Generation

* `multi_shots`: Optional. Enable multi-shot generation. Use `true` or `false`. Default: `false`.

### Sound

* `sound`: Optional. Set `true` to generate audio or `false` for a silent single shot. Multiple shots require `true`. Default: `true`.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `16:9`, `9:16`. Default: `1:1`.


# Kling 3.0 Turbo Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-turbo-pro-image-to-video

api-manual/video-series/kwaivgi-kling-v3-turbo-pro-image-to-video.json POST /api/generate/submit
Bring reference images into motion with Kling 3.0 Turbo Pro, using text and shot prompts to guide a video sequence.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Bring reference images into motion with Kling 3.0 Turbo Pro, using text and shot prompts to guide a video sequence.

## Models

* **kwaivgi/kling-v3-turbo-pro/image-to-video** - Generate video from image input.

## Features

* Animate a single reference image.
* Direct one scene with `prompt`, or use `multi_prompt` for up to six shots.
* The input image determines the output aspect ratio.

## Duration Options

* **3-15 seconds** - Accepted range for an explicit `duration`. Without `multi_prompt`, the default is **5 seconds**.
* **Multi-shot output** - Each shot accepts `1`-`15` seconds and defaults to `5`. The total must not exceed **15 seconds**.
* With `multi_prompt`, omit `duration` to use the shot total, including totals of **1 or 2 seconds**. If supplied, `duration` must equal the shot total and remain within `3`-`15` seconds.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt` is optional and supports up to **2,500 characters**. Blank optional prompts are treated as omitted.
* Use either `prompt` or `multi_prompt`; do not combine them.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* Images must be publicly accessible and directly downloadable.

### Multi-Prompt

* `multi_prompt` accepts **1-6 shot objects** and cannot be combined with `prompt`.
* Each shot requires a nonblank `prompt` of at most **2,500 characters**.
* Each shot can include an integer `duration` from **1 to 15 seconds**; the default is **5 seconds**.
* The combined shot duration must not exceed **15 seconds**. See Duration Options for the top-level `duration` rules.

### Aspect Ratio

* The output aspect ratio follows the input image.
* `aspect_ratio` is ignored by this endpoint and can be omitted.

## Pricing

With `multi_prompt`, billing uses the combined shot duration. Otherwise, it uses the requested `duration` or its default.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 3.0 Turbo Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-turbo-pro-text-to-video

api-manual/video-series/kwaivgi-kling-v3-turbo-pro-text-to-video.json POST /api/generate/submit
Build video scenes with Kling 3.0 Turbo Pro, using text and individual shot prompts to direct action and composition.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Build video scenes with Kling 3.0 Turbo Pro, using text and individual shot prompts to direct action and composition.

## Models

* **kwaivgi/kling-v3-turbo-pro/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Direct one scene with `prompt`, or use `multi_prompt` for up to six shots.

## Duration Options

* **3-15 seconds** - Accepted range for an explicit `duration`. Without `multi_prompt`, the default is **5 seconds**.
* **Multi-shot output** - Each shot accepts `1`-`15` seconds and defaults to `5`. The total must not exceed **15 seconds**.
* With `multi_prompt`, omit `duration` to use the shot total, including totals of **1 or 2 seconds**. If supplied, `duration` must equal the shot total and remain within `3`-`15` seconds.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt` is optional and supports up to **2,500 characters**. Blank optional prompts are treated as omitted.
* Use either `prompt` or `multi_prompt`; do not combine them.

### Multi-Prompt

* `multi_prompt` accepts **1-6 shot objects** and cannot be combined with `prompt`.
* Each shot requires a nonblank `prompt` of at most **2,500 characters**.
* Each shot can include an integer `duration` from **1 to 15 seconds**; the default is **5 seconds**.
* The combined shot duration must not exceed **15 seconds**. See Duration Options for the top-level `duration` rules.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`, `1:1`. Default: `16:9`.

## Pricing

With `multi_prompt`, billing uses the combined shot duration. Otherwise, it uses the requested `duration` or its default.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 3.0 Turbo Standard Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-turbo-std-image-to-video

api-manual/video-series/kwaivgi-kling-v3-turbo-std-image-to-video.json POST /api/generate/submit
Bring reference images into motion with Kling 3.0 Turbo Standard, using text and shot prompts to guide a video sequence.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Bring reference images into motion with Kling 3.0 Turbo Standard, using text and shot prompts to guide a video sequence.

## Models

* **kwaivgi/kling-v3-turbo-std/image-to-video** - Generate video from image input.

## Features

* Animate a single reference image.
* Direct one scene with `prompt`, or use `multi_prompt` for up to six shots.
* The input image determines the output aspect ratio.

## Duration Options

* **3-15 seconds** - Accepted range for an explicit `duration`. Without `multi_prompt`, the default is **5 seconds**.
* **Multi-shot output** - Each shot accepts `1`-`15` seconds and defaults to `5`. The total must not exceed **15 seconds**.
* With `multi_prompt`, omit `duration` to use the shot total, including totals of **1 or 2 seconds**. If supplied, `duration` must equal the shot total and remain within `3`-`15` seconds.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt` is optional and supports up to **2,500 characters**. Blank optional prompts are treated as omitted.
* Use either `prompt` or `multi_prompt`; do not combine them.

### Image URLs

* `image_urls`: Required. Provide exactly 1 image URL.
* Images must be publicly accessible and directly downloadable.

### Multi-Prompt

* `multi_prompt` accepts **1-6 shot objects** and cannot be combined with `prompt`.
* Each shot requires a nonblank `prompt` of at most **2,500 characters**.
* Each shot can include an integer `duration` from **1 to 15 seconds**; the default is **5 seconds**.
* The combined shot duration must not exceed **15 seconds**. See Duration Options for the top-level `duration` rules.

### Aspect Ratio

* The output aspect ratio follows the input image.
* `aspect_ratio` is ignored by this endpoint and can be omitted.

## Pricing

With `multi_prompt`, billing uses the combined shot duration. Otherwise, it uses the requested `duration` or its default.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 3.0 Turbo Standard Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v3-turbo-std-text-to-video

api-manual/video-series/kwaivgi-kling-v3-turbo-std-text-to-video.json POST /api/generate/submit
Build video scenes with Kling 3.0 Turbo Standard, using text and individual shot prompts to direct action and composition.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Build video scenes with Kling 3.0 Turbo Standard, using text and individual shot prompts to direct action and composition.

## Models

* **kwaivgi/kling-v3-turbo-std/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Direct one scene with `prompt`, or use `multi_prompt` for up to six shots.

## Duration Options

* **3-15 seconds** - Accepted range for an explicit `duration`. Without `multi_prompt`, the default is **5 seconds**.
* **Multi-shot output** - Each shot accepts `1`-`15` seconds and defaults to `5`. The total must not exceed **15 seconds**.
* With `multi_prompt`, omit `duration` to use the shot total, including totals of **1 or 2 seconds**. If supplied, `duration` must equal the shot total and remain within `3`-`15` seconds.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place the parameters below inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt` is optional and supports up to **2,500 characters**. Blank optional prompts are treated as omitted.
* Use either `prompt` or `multi_prompt`; do not combine them.

### Multi-Prompt

* `multi_prompt` accepts **1-6 shot objects** and cannot be combined with `prompt`.
* Each shot requires a nonblank `prompt` of at most **2,500 characters**.
* Each shot can include an integer `duration` from **1 to 15 seconds**; the default is **5 seconds**.
* The combined shot duration must not exceed **15 seconds**. See Duration Options for the top-level `duration` rules.

### Aspect Ratio

* `aspect_ratio`: `16:9`, `9:16`, `1:1`. Default: `16:9`.

## Pricing

With `multi_prompt`, billing uses the combined shot duration. Otherwise, it uses the requested `duration` or its default.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 4.0 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v4-0-image-to-video

api-manual/video-series/kwaivgi-kling-v4-0-image-to-video.json POST /api/generate/submit
The Kling 4.0 Image to Video API animates a start frame, optionally ending on a second image, into a 3–30 second 720p or 1080p video.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Kling 4.0 Image to Video animates a start frame into a 3–30 second video at 720p or 1080p. Add a second image and Kling 4.0 uses it as the end frame, generating the motion between the two. It suits product shots, character animation and before/after transitions.

Kling 4.0 is Kuaishou's Kling AI video model. Vidgo exposes it as four model IDs: [Text to Video](/api-manual/video-series/kwaivgi-kling-v4-0-text-to-video), [Image to Video](/api-manual/video-series/kwaivgi-kling-v4-0-image-to-video), [Reference to Video](/api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video) and [Video Edit](/api-manual/video-series/kwaivgi-kling-v4-0-video-edit).

## Models

* **kwaivgi/kling-v4.0/image-to-video** — Image-to-video with Kling 4.0 from a start frame and an optional end frame.

## Features

* Image-to-video up to 30 seconds from one start frame.
* First and last frame control with a second image.
* Single or multi-shot prompts.
* 720p or 1080p output; the frame shape follows the input image.
* Output includes an audio track.

## Input modes

* **Start frame**: `image_urls` with one image.
* **Start and end frame**: `image_urls` with two images, start frame first.

Either mode takes a single `prompt` or a `multi_prompt` shot list:

* **Single prompt**: send `prompt` and, optionally, `duration`.
* **Multi-shot**: send `multi_prompt` with one entry per shot. The clip length is the sum of the shot durations; leave `duration` out or set it to that sum.

## Required parameters

* `image_urls` (string\[]): One or two public HTTP(S) image URLs. The first is the start frame; the optional second is the end frame.
* `prompt` (string) **or** `multi_prompt` (object\[]): send exactly one.
  * `prompt`: 1–26,000 characters after trimming.
  * `multi_prompt`: a non-empty list of shots `{prompt, duration?}`. Each shot `prompt` is 1–512 characters; each shot `duration` is an integer `1`–`30` (default `5`). The shot durations must total 3–30 seconds.

## Optional parameters

* `duration` (integer): Output length, `3`–`30` seconds. Default `5`; with `multi_prompt` the default is the shot total, and an explicit value must equal it.
* `resolution` (string): `720p` or `1080p`. Default `720p`.

## Output

The finished task returns one MP4 video, with an audio track, in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Start frame only**: one image of a dancer with a motion and music prompt, 5 seconds at 720p.
* **Start and end frame, 1080p, with callback**: a bud and an open flower as start and end frames, an 8-second time-lapse and a `callback_url`.

## Notes

* `aspect_ratio` is not supported on this ID because the input image sets the frame; sending it returns 400.
* To guide the video with extra reference images, reference videos or subjects, use [Kling 4.0 Reference to Video](/api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video).
* Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and the output `duration`. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* Output is 720p or 1080p only; there is no 4K or HDR tier, no audio on/off switch and no `seed` on this model.


# Kling 4.0 Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video

api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video.json POST /api/generate/submit
The Kling 4.0 Reference to Video API guides 3–30 second videos with up to 10 reference images, 5 reference videos and 7 elements, cited as @Image1 or @Element1.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Kling 4.0 Reference to Video generates a 3–30 second video guided by reference images, reference videos and reusable subjects (elements), so characters, products and settings stay consistent. Cite each reference in the prompt by its position, such as `@Image1`, `@Video1` or `@Element1`.

Kling 4.0 is Kuaishou's Kling AI video model. Vidgo exposes it as four model IDs: [Text to Video](/api-manual/video-series/kwaivgi-kling-v4-0-text-to-video), [Image to Video](/api-manual/video-series/kwaivgi-kling-v4-0-image-to-video), [Reference to Video](/api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video) and [Video Edit](/api-manual/video-series/kwaivgi-kling-v4-0-video-edit).

## Models

* **kwaivgi/kling-v4.0/reference-to-video** — Reference-to-video with Kling 4.0 from images, videos and elements, 3–30 seconds, 720p or 1080p.

## Features

* Up to 10 reference images, 5 reference videos and 7 elements, 15 assets in total.
* Elements built from 2–4 images or one video of the same subject.
* Optional start frame via `image_urls`.
* Single or multi-shot prompts; 3–30 seconds at 720p or 1080p.
* Five aspect ratios: `auto`, `16:9`, `9:16`, `1:1`, `21:9`.

## Input modes

Send at least one reference input; you can combine them:

* **Reference images** (`reference_image_urls`): cite as `@Image1`, `@Image2`, ….
* **Reference videos** (`reference_video_urls`): cite as `@Video1`, `@Video2`, …. Billed as input video.
* **Elements** (`kling_elements`): cite as `@Element1`, `@Element2`, …. Video-based elements are billed as input video.

Add `image_urls` with one image to fix the opening frame. Then choose a single `prompt` or a `multi_prompt` shot list:

* **Single prompt**: send `prompt` and, optionally, `duration`.
* **Multi-shot**: send `multi_prompt` with one entry per shot. The clip length is the sum of the shot durations; leave `duration` out or set it to that sum.

## Required parameters

* At least one of `reference_image_urls`, `reference_video_urls` or `kling_elements`.
* `prompt` (string) **or** `multi_prompt` (object\[]): send exactly one.
  * `prompt`: 1–26,000 characters after trimming.
  * `multi_prompt`: a non-empty list of shots `{prompt, duration?}`. Each shot `prompt` is 1–512 characters; each shot `duration` is an integer `1`–`30` (default `5`). The shot durations must total 3–30 seconds.

## Optional parameters

* `reference_image_urls` (string\[]): Up to 10 public HTTP(S) image URLs.
* `reference_video_urls` (string\[]): Up to 5 public HTTP(S) video URLs, each 3–30 seconds.
* `kling_elements` (object\[]): Up to 7 elements, at most 3 video-based. Each element has exactly one of:
  * `element_input_urls` (string\[]): 2–4 image URLs of the subject.
  * `element_input_video_urls` (string\[]): exactly 1 video URL of the subject (3–30 seconds).
* `image_urls` (string\[]): One start-frame image URL.
* `duration` (integer): Output length, `3`–`30` seconds. Default `5`; with `multi_prompt` the default is the shot total, and an explicit value must equal it.
* `resolution` (string): `720p` or `1080p`. Default `720p`.
* `aspect_ratio` (string): `auto`, `16:9`, `9:16`, `1:1` or `21:9`. Default `auto`.

## Output

The finished task returns one MP4 video, with an audio track, in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Reference images**: a character and a cafe image, cited as `@Image1` and `@Image2`, 8 seconds at 720p.
* **Elements plus a reference video, with callback**: one robot element from two images and a dance clip cited as `@Video1`, 10 seconds at 1080p, `9:16`.

## Notes

* Limits: start frame plus reference images ≤ 10; all assets (start frame, reference images, reference videos, elements) ≤ 15; video-based elements ≤ 3.
* Input videos are downloaded and measured server-side (up to 100 MB each). Each must run 3–30 seconds, and all input videos together must not exceed 30 seconds.
* `voice_id` and other extra fields inside an element return 400.
* Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and duration. When an input video is present, a separate input-video rate applies to billed seconds = output duration + the larger of the output duration and the total input video length (summed, then rounded up to whole seconds). A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates. Requests with reference images or image elements only use the standard rate on the output duration.
* Output is 720p or 1080p only; there is no 4K or HDR tier, no audio on/off switch and no `seed` on this model.


# Kling 4.0 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v4-0-text-to-video

api-manual/video-series/kwaivgi-kling-v4-0-text-to-video.json POST /api/generate/submit
The Kling 4.0 Text to Video API generates 3–30 second videos from a prompt or multi-shot prompts at 720p or 1080p, with stereo audio and five aspect ratios.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Kling 4.0 Text to Video generates a 3–30 second video from a single prompt or from a multi-shot script, at 720p or 1080p, with an audio track in the output. Use it for single-take ads, short-drama scenes and social clips when you have no source image.

Kling 4.0 is Kuaishou's Kling AI video model. Vidgo exposes it as four model IDs: [Text to Video](/api-manual/video-series/kwaivgi-kling-v4-0-text-to-video), [Image to Video](/api-manual/video-series/kwaivgi-kling-v4-0-image-to-video), [Reference to Video](/api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video) and [Video Edit](/api-manual/video-series/kwaivgi-kling-v4-0-video-edit).

## Models

* **kwaivgi/kling-v4.0/text-to-video** — Text-to-video with Kling 4.0, 3–30 seconds, single or multi-shot prompts, 720p or 1080p.

## Features

* Text-to-video up to 30 seconds in one pass.
* Multi-shot prompts with per-shot durations via `multi_prompt`.
* 720p or 1080p output.
* Five aspect ratios: `auto`, `16:9`, `9:16`, `1:1`, `21:9`.
* Output includes an audio track; describe dialogue, sound effects or ambience in the prompt.

## Input modes

* **Single prompt**: send `prompt` and, optionally, `duration`.
* **Multi-shot**: send `multi_prompt` with one entry per shot. The clip length is the sum of the shot durations; leave `duration` out or set it to that sum.

## Required parameters

* `prompt` (string) **or** `multi_prompt` (object\[]): send exactly one.
  * `prompt`: 1–26,000 characters after trimming.
  * `multi_prompt`: a non-empty list of shots `{prompt, duration?}`. Each shot `prompt` is 1–512 characters; each shot `duration` is an integer `1`–`30` (default `5`). The shot durations must total 3–30 seconds.

## Optional parameters

* `duration` (integer): Output length, `3`–`30` seconds. Default `5`; with `multi_prompt` the default is the shot total, and an explicit value must equal it.
* `resolution` (string): `720p` or `1080p`. Default `720p`.
* `aspect_ratio` (string): `auto`, `16:9`, `9:16`, `1:1` or `21:9`. Default `auto`.

## Output

The finished task returns one MP4 video, with an audio track, in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Single prompt, 10 seconds**: one `prompt` with scene, camera and sound cues, `16:9`, 720p.
* **Three-shot 1080p sequence with callback**: `multi_prompt` with shots of 4, 6 and 5 seconds (15 seconds in total), `9:16`, and a `callback_url`.

## Notes

* Sending both `prompt` and `multi_prompt`, or neither, returns 400. With `multi_prompt`, a `duration` that differs from the shot total returns 400.
* This ID accepts text only. To start from an image, use [Kling 4.0 Image to Video](/api-manual/video-series/kwaivgi-kling-v4-0-image-to-video); to guide the video with images, videos or subjects, use [Kling 4.0 Reference to Video](/api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video).
* Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and the output `duration`. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* Output is 720p or 1080p only; there is no 4K or HDR tier, no audio on/off switch and no `seed` on this model.


# Kling 4.0 Video Edit
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-v4-0-video-edit

api-manual/video-series/kwaivgi-kling-v4-0-video-edit.json POST /api/generate/submit
The Kling 4.0 Video Edit API edits a 3–30 second source video from a text instruction, with optional reference images and subjects, at 720p or 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Kling 4.0 Video Edit changes an existing 3–30 second video from a text instruction: restyle the scene, swap objects or backgrounds, or adjust the action, while keeping the source motion. Reference images and image-based subjects can show exactly what to insert.

Kling 4.0 is Kuaishou's Kling AI video model. Vidgo exposes it as four model IDs: [Text to Video](/api-manual/video-series/kwaivgi-kling-v4-0-text-to-video), [Image to Video](/api-manual/video-series/kwaivgi-kling-v4-0-image-to-video), [Reference to Video](/api-manual/video-series/kwaivgi-kling-v4-0-reference-to-video) and [Video Edit](/api-manual/video-series/kwaivgi-kling-v4-0-video-edit).

## Models

* **kwaivgi/kling-v4.0/video-edit** — Video edit with Kling 4.0: text-guided editing of a 3–30 second source video, 720p or 1080p.

## Features

* Text-guided video edit of a 3–30 second source clip.
* Optional reference images (`@Image1`, …) and image-based elements (`@Element1`, …).
* Output length matches the source by default, or set `duration`.
* 720p or 1080p output.

## Required parameters

* `video_url` (string): Public HTTP(S) URL of the source video, 3–30 seconds, up to 100 MB.
* `prompt` (string): The edit instruction, 1–26,000 characters after trimming.

## Optional parameters

* `duration` (integer): Output length, `3`–`30` seconds. No default: if omitted, the output matches the source.
* `resolution` (string): `720p` or `1080p`. Default `720p`.
* `reference_image_urls` (string\[]): Up to 10 public HTTP(S) image URLs.
* `kling_elements` (object\[]): Up to 7 image-based elements, each `{element_input_urls}` with 2–4 image URLs. Video elements are not supported here.

## Output

The finished task returns one MP4 video, with an audio track, in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Restyle, keep source length**: a winter makeover of a street clip without `duration`, so the output follows the source.
* **Replace an object with a reference image, with callback**: swap a phone for the product in `@Image1`, 10 seconds at 1080p, with a `callback_url`.

## Notes

* `multi_prompt`, `image_urls`, `reference_video_urls`, `aspect_ratio` and video-based elements are not supported on this ID and return 400.
* The source video is downloaded and measured server-side; it must run 3–30 seconds and be at most 100 MB.
* Fields not listed above return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and duration. When an input video is present, a separate input-video rate applies to billed seconds = output duration + the larger of the output duration and the total input video length (summed, then rounded up to whole seconds). A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates. Without `duration`, the output duration is billed as the source length rounded up.
* Output is 720p or 1080p only; there is no 4K or HDR tier, no audio on/off switch and no `seed` on this model.


# Kling Avatar 2.0 Pro
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-ai-avatar-v2-pro

api-manual/video-series/kwaivgi-kling-video-ai-avatar-v2-pro.json POST /api/generate/submit
Create a speaking avatar with Kling Avatar 2.0 Pro, combining a character image and an audio performance with optional text direction.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create a speaking avatar with Kling Avatar 2.0 Pro, combining a character image and an audio performance with optional text direction.

## Models

* **kwaivgi/kling-video/ai-avatar/v2/pro** - Animate a character image with driving audio.

## Features

* Animate a character image with driving audio.
* Accepts exactly 1 image in `image_urls`.

## Duration Options

* **Driving audio: 2-60 seconds** - The service detects the duration of `audio_url`.
* The audio must be publicly reachable over HTTP(S) and no larger than **5 MiB**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials. Image inputs also accept image Data URIs or raw Base64. Image conversion and media validation complete before credits are deducted.

### Image URLs

* `image_urls`: Required. Supply exactly 1 item. Each item may be a public HTTP(S) URL, image Data URI, or raw Base64. Image order is preserved.

### Audio URL

* `audio_url`: Required. Provide the input audio URL. Use a publicly accessible HTTP(S) URL.

### Prompt

* `prompt`: Optional. Describe the scene and desired motion. Maximum **2500 characters** after trimming.
* Leading and trailing whitespace is trimmed. Omitted, null, empty, and whitespace-only values are treated as omitted.

## Pricing

Usage is based on the detected driving-audio duration, rounded up to a whole second.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling Avatar 2.0 Standard
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-ai-avatar-v2-standard

api-manual/video-series/kwaivgi-kling-video-ai-avatar-v2-standard.json POST /api/generate/submit
Create a speaking avatar with Kling Avatar 2.0 Standard, combining a character image and an audio performance with optional text direction.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create a speaking avatar with Kling Avatar 2.0 Standard, combining a character image and an audio performance with optional text direction.

## Models

* **kwaivgi/kling-video/ai-avatar/v2/standard** - Animate a character image with driving audio.

## Features

* Animate a character image with driving audio.
* Accepts exactly 1 image in `image_urls`.

## Duration Options

* **Driving audio: 2-60 seconds** - The service detects the duration of `audio_url`.
* The audio must be publicly reachable over HTTP(S) and no larger than **5 MiB**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials. Image inputs also accept image Data URIs or raw Base64. Image conversion and media validation complete before credits are deducted.

### Image URLs

* `image_urls`: Required. Supply exactly 1 item. Each item may be a public HTTP(S) URL, image Data URI, or raw Base64. Image order is preserved.

### Audio URL

* `audio_url`: Required. Provide the input audio URL. Use a publicly accessible HTTP(S) URL.

### Prompt

* `prompt`: Optional. Describe the scene and desired motion. Maximum **2500 characters** after trimming.
* Leading and trailing whitespace is trimmed. Omitted, null, empty, and whitespace-only values are treated as omitted.

## Pricing

Usage is based on the detected driving-audio duration, rounded up to a whole second.

Check the Vidgo dashboard for the current rate.

Credits are deducted on submission and refunded if the task fails.


# Kling O3 4K Image To Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-o3-4k-image-to-video

api-manual/video-series/kwaivgi-kling-video-o3-4k-image-to-video.json POST /api/generate/submit
Generate 3–15-second videos with Kling O3 4K from a starting image and an optional ending frame. Outputs 4K video. Supports single-shot and multi-shot generation. Explicitly set sound; multi-shot requires audio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Parameters

Use model `kwaivgi/kling-video-o3-4k/image-to-video`. Put generation parameters inside `input`.

* **`image_urls`**: Required: an array of 1–2 images. The first is the starting frame; the optional second is the ending frame. Accepts public HTTP(S) URLs, image Data URIs or raw Base64 image data.
* **`prompt`**: Required when multi\_shots=false. Describe the scene, action and camera movement. Use nonblank text, up to 2,500 characters after trimming. Omit prompt when multi\_shots=true.
* **`multi_prompt`**: Required when multi\_shots=true; omit in single-shot mode. Supply at least one shot, each with a nonblank prompt of up to 2,500 characters and an integer duration of 1–12 seconds. Shot durations must sum to the top-level duration (3–15 seconds). Extra fields inside a shot are ignored.
* **`duration`**: Required integer from 3 to 15 seconds. In multi-shot mode, this must equal the sum of all shot durations. Credits are calculated using this value.
* **`multi_shots`**: Required: explicitly send false for a single shot or true for multiple shots. Single-shot mode requires prompt. Multi-shot mode requires multi\_prompt and sound=true, with no nonblank top-level prompt. Omitting this field is an error.
* **`sound`**: Required: explicitly send true to generate audio or false for video without audio. Single-shot mode accepts either value; multi-shot mode requires true.
* **`aspect_ratio`**: May be omitted. If provided, use 16:9, 9:16 or 1:1. Image-to-video ignores this value; framing is determined by the input image.

Use standard JSON numbers and booleans. For compatibility, integer strings such as "5" are accepted. Boolean strings true/1/yes/y/on mean true; false/0/no/n/off mean false. These strings are case-insensitive and trimmed. Numeric 1 and 0 are also accepted for boolean fields. Numeric and boolean prompt values are converted to text; 0, false and null are treated as empty. Objects and arrays are not accepted as prompts. Unsupported input fields are rejected. HTTP(S) URLs must have a hostname and no embedded credentials.

## Billing Factors

Billed per generated second using the 4K model tier. Video duration (`duration`) determines usage; enabling or disabling audio (`sound`) does not change the rate.

Failed generation tasks are refunded.

## Default example: The Greenhouse Trap

This is the same default example shown on the [model page](https://vidgo.ai/apis/models/kwaivgi/kling-video-o3-4k/image-to-video).

```json theme={null}
{
  "model": "kwaivgi/kling-video-o3-4k/image-to-video",
  "input": {
    "duration": 4,
    "sound": true,
    "multi_shots": false,
    "prompt": "Begin with the supplied botanical macro. The foreground Venus flytrap slowly closes its two lobes around their central hinge, ending with marginal cilia interlaced. Preserve its vein pattern, dew beads, surrounding leaves and blurred greenhouse glazing. Droplet highlights shift naturally. Fixed macro camera, realistic plant motion; no insect, animal features or new leaves. Audio: quiet greenhouse ambience and one distant drip; no music. No text, logos or watermarks.",
    "image_urls": [
      "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-4k/image-to-video/v1/01/input-start.png"
    ]
  }
}
```

![Starting frame](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-4k/image-to-video/v1/01/input-start.png)

<video />

[Download this video](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-4k/image-to-video/v1/01/output.mp4)

### Completed result from the status endpoint

`GET /api/generate/status/{task_id}`

The response below records this verified example. For a new generation, query the task ID returned by your own submission.

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "MQX9Q1HSIUMLPP44",
    "status": "finished",
    "files": [
      {
        "file_type": "video",
        "file_url": "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-4k/image-to-video/v1/01/output.mp4"
      }
    ],
    "created_time": "2026-09-22T18:50:38",
    "error_message": null,
    "progress": 100
  }
}
```

> [!NOTE]
>
> The API validates prompts up to 2,500 characters. Some multi-shot generation requests have failed when a shot prompt exceeded 512 characters. We recommend keeping each shot prompt within 512 characters; this recommendation does not change the API validation limit.


# Kling O3 4K Text To Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-o3-4k-text-to-video

api-manual/video-series/kwaivgi-kling-video-o3-4k-text-to-video.json POST /api/generate/submit
Generate 3–15-second videos from text prompts with Kling O3 4K. Outputs 4K video. Supports single-shot and multi-shot generation. Explicitly set sound; multi-shot requires audio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Parameters

Use model `kwaivgi/kling-video-o3-4k/text-to-video`. Put generation parameters inside `input`.

* **`prompt`**: Required when multi\_shots=false. Describe the scene, action and camera movement. Use nonblank text, up to 2,500 characters after trimming. Omit prompt when multi\_shots=true.
* **`multi_prompt`**: Required when multi\_shots=true; omit in single-shot mode. Supply at least one shot, each with a nonblank prompt of up to 2,500 characters and an integer duration of 1–12 seconds. Shot durations must sum to the top-level duration (3–15 seconds). Extra fields inside a shot are ignored.
* **`duration`**: Required integer from 3 to 15 seconds. In multi-shot mode, this must equal the sum of all shot durations. Credits are calculated using this value.
* **`multi_shots`**: Required: explicitly send false for a single shot or true for multiple shots. Single-shot mode requires prompt. Multi-shot mode requires multi\_prompt and sound=true, with no nonblank top-level prompt. Omitting this field is an error.
* **`sound`**: Required: explicitly send true to generate audio or false for video without audio. Single-shot mode accepts either value; multi-shot mode requires true.
* **`aspect_ratio`**: Optional: 16:9 (landscape), 9:16 (portrait) or 1:1 (square). Sets the video aspect ratio.

Use standard JSON numbers and booleans. For compatibility, integer strings such as "5" are accepted. Boolean strings true/1/yes/y/on mean true; false/0/no/n/off mean false. These strings are case-insensitive and trimmed. Numeric 1 and 0 are also accepted for boolean fields. Numeric and boolean prompt values are converted to text; 0, false and null are treated as empty. Objects and arrays are not accepted as prompts. Unsupported input fields are rejected.

## Billing Factors

Billed per generated second using the 4K model tier. Video duration (`duration`) determines usage; enabling or disabling audio (`sound`) does not change the rate.

Failed generation tasks are refunded.

## Default example: Rain Along the Stone Gallery

This is the same default example shown on the [model page](https://vidgo.ai/apis/models/kwaivgi/kling-video-o3-4k/text-to-video).

```json theme={null}
{
  "model": "kwaivgi/kling-video-o3-4k/text-to-video",
  "input": {
    "duration": 4,
    "sound": true,
    "multi_shots": false,
    "prompt": "Photoreal slow lateral dolly through an empty ancient temple gallery in rain. Weathered floral stone reliefs fill the foreground; rows of pillars recede toward a quiet courtyard. Water follows carved grooves and drips from worn edges, revealing mineral grains, chisel marks and moss in cracks. Soft overcast daylight, subtle wet highlights, stable architecture and rich fine detail. Audio: gentle rain and isolated drips; no music. No text, logos or watermarks.",
    "aspect_ratio": "16:9"
  }
}
```

<video />

[Download this video](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-4k/text-to-video/v1/01/output.mp4)

### Completed result from the status endpoint

`GET /api/generate/status/{task_id}`

The response below records this verified example. For a new generation, query the task ID returned by your own submission.

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "EE4XVIYVL5IAX4YG",
    "status": "finished",
    "files": [
      {
        "file_type": "video",
        "file_url": "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-4k/text-to-video/v1/01/output.mp4"
      }
    ],
    "created_time": "2026-09-22T18:48:45",
    "error_message": null,
    "progress": 100
  }
}
```

> [!NOTE]
>
> The API validates prompts up to 2,500 characters. Some multi-shot generation requests have failed when a shot prompt exceeded 512 characters. We recommend keeping each shot prompt within 512 characters; this recommendation does not change the API validation limit.


# Kling O3 Pro Image To Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-o3-pro-image-to-video

api-manual/video-series/kwaivgi-kling-video-o3-pro-image-to-video.json POST /api/generate/submit
Generate 3–15-second videos with Kling O3 Pro from a starting image and an optional ending frame. Supports single-shot and multi-shot generation. Explicitly set sound; multi-shot requires audio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Parameters

Use model `kwaivgi/kling-video-o3-pro/image-to-video`. Put generation parameters inside `input`.

* **`image_urls`**: Required: an array of 1–2 images. The first is the starting frame; the optional second is the ending frame. Accepts public HTTP(S) URLs, image Data URIs or raw Base64 image data.
* **`prompt`**: Required when multi\_shots=false. Describe the scene, action and camera movement. Use nonblank text, up to 2,500 characters after trimming. Omit prompt when multi\_shots=true.
* **`multi_prompt`**: Required when multi\_shots=true; omit in single-shot mode. Supply at least one shot, each with a nonblank prompt of up to 2,500 characters and an integer duration of 1–12 seconds. Shot durations must sum to the top-level duration (3–15 seconds). Extra fields inside a shot are ignored.
* **`duration`**: Required integer from 3 to 15 seconds. In multi-shot mode, this must equal the sum of all shot durations. Credits are calculated using this value.
* **`multi_shots`**: Required: explicitly send false for a single shot or true for multiple shots. Single-shot mode requires prompt. Multi-shot mode requires multi\_prompt and sound=true, with no nonblank top-level prompt. Omitting this field is an error.
* **`sound`**: Required: explicitly send true to generate audio or false for video without audio. Single-shot mode accepts either value; multi-shot mode requires true.
* **`aspect_ratio`**: May be omitted. If provided, use 16:9, 9:16 or 1:1. Image-to-video ignores this value; framing is determined by the input image.

Use standard JSON numbers and booleans. For compatibility, integer strings such as "5" are accepted. Boolean strings true/1/yes/y/on mean true; false/0/no/n/off mean false. These strings are case-insensitive and trimmed. Numeric 1 and 0 are also accepted for boolean fields. Numeric and boolean prompt values are converted to text; 0, false and null are treated as empty. Objects and arrays are not accepted as prompts. Unsupported input fields are rejected. HTTP(S) URLs must have a hostname and no embedded credentials.

## Billing Factors

Billing depends on the model tier, video duration (`duration`), and whether audio is generated (`sound`). Usage is charged per generated second.

Failed generation tasks are refunded.

## Default example: At the Volcano Observatory

This is the same default example shown on the [model page](https://vidgo.ai/apis/models/kwaivgi/kling-video-o3-pro/image-to-video).

```json theme={null}
{
  "model": "kwaivgi/kling-video-o3-pro/image-to-video",
  "input": {
    "duration": 5,
    "sound": true,
    "multi_shots": true,
    "multi_prompt": [
      {
        "prompt": "Begin with the supplied photograph. Preserve the female geologist, white helmet, charcoal jacket with teal collar, orange instrument and safety railing. Medium-wide view: she watches a distant lava fountain beyond the secured platform. Warm lava light touches her jacket; wind moves a loose hair strand. Restrained documentary realism, safe distance. Audio: quiet wind and low distant rumble. No text, logos or watermarks.",
        "duration": 2
      },
      {
        "prompt": "Cut to a closer three-quarter view of the same geologist on the same platform. Preserve her face, white helmet, teal collar and orange instrument. She lowers her gaze from the volcano to the instrument and makes one small thumb adjustment. Maintain warm lava reflections against cool twilight. Finish looking down. Low distant rumble and glove movement; no speech, alarm or music. No text, logos or watermarks.",
        "duration": 3
      }
    ],
    "image_urls": [
      "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-pro/image-to-video/v1/01/input-start.png"
    ]
  }
}
```

![Starting frame](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-pro/image-to-video/v1/01/input-start.png)

<video />

[Download this video](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-pro/image-to-video/v1/01/output.mp4)

### Completed result from the status endpoint

`GET /api/generate/status/{task_id}`

The response below records this verified example. For a new generation, query the task ID returned by your own submission.

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "WTSYQAI2J76PVHFW",
    "status": "finished",
    "files": [
      {
        "file_type": "video",
        "file_url": "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-pro/image-to-video/v1/01/output.mp4"
      }
    ],
    "created_time": "2026-09-22T18:46:21",
    "error_message": null,
    "progress": 100
  }
}
```

> [!NOTE]
>
> The API validates prompts up to 2,500 characters. Some multi-shot generation requests have failed when a shot prompt exceeded 512 characters. We recommend keeping each shot prompt within 512 characters; this recommendation does not change the API validation limit.


# Kling O3 Pro Text To Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-o3-pro-text-to-video

api-manual/video-series/kwaivgi-kling-video-o3-pro-text-to-video.json POST /api/generate/submit
Generate 3–15-second videos from text prompts with Kling O3 Pro. Supports single-shot and multi-shot generation. Explicitly set sound; multi-shot requires audio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Parameters

Use model `kwaivgi/kling-video-o3-pro/text-to-video`. Put generation parameters inside `input`.

* **`prompt`**: Required when multi\_shots=false. Describe the scene, action and camera movement. Use nonblank text, up to 2,500 characters after trimming. Omit prompt when multi\_shots=true.
* **`multi_prompt`**: Required when multi\_shots=true; omit in single-shot mode. Supply at least one shot, each with a nonblank prompt of up to 2,500 characters and an integer duration of 1–12 seconds. Shot durations must sum to the top-level duration (3–15 seconds). Extra fields inside a shot are ignored.
* **`duration`**: Required integer from 3 to 15 seconds. In multi-shot mode, this must equal the sum of all shot durations. Credits are calculated using this value.
* **`multi_shots`**: Required: explicitly send false for a single shot or true for multiple shots. Single-shot mode requires prompt. Multi-shot mode requires multi\_prompt and sound=true, with no nonblank top-level prompt. Omitting this field is an error.
* **`sound`**: Required: explicitly send true to generate audio or false for video without audio. Single-shot mode accepts either value; multi-shot mode requires true.
* **`aspect_ratio`**: Optional: 16:9 (landscape), 9:16 (portrait) or 1:1 (square). Sets the video aspect ratio.

Use standard JSON numbers and booleans. For compatibility, integer strings such as "5" are accepted. Boolean strings true/1/yes/y/on mean true; false/0/no/n/off mean false. These strings are case-insensitive and trimmed. Numeric 1 and 0 are also accepted for boolean fields. Numeric and boolean prompt values are converted to text; 0, false and null are treated as empty. Objects and arrays are not accepted as prompts. Unsupported input fields are rejected.

## Billing Factors

Billing depends on the model tier, video duration (`duration`), and whether audio is generated (`sound`). Usage is charged per generated second.

Failed generation tasks are refunded.

## Default example: A Quiet Apology

This is the same default example shown on the [model page](https://vidgo.ai/apis/models/kwaivgi/kling-video-o3-pro/text-to-video).

```json theme={null}
{
  "model": "kwaivgi/kling-video-o3-pro/text-to-video",
  "input": {
    "duration": 5,
    "sound": true,
    "multi_shots": true,
    "multi_prompt": [
      {
        "prompt": "Cinematic whimsical realism in a warm oak library. Medium two-shot: a small ivory robot with an oval face and amber eyes shelves a blue book, accidentally knocking one red book onto the floor. Beside it, an adult librarian wears a green cardigan and round glasses. Both remain visible under warm reading lamps. One distinct book thud against quiet room tone. Unmarked book covers. No text, logos or watermarks.",
        "duration": 2
      },
      {
        "prompt": "Cut closer to the same ivory robot and green-cardigan librarian in the same oak library. The librarian raises one finger to her lips. The robot tilts its head apologetically and softly says exactly \"Sorry.\" in a gentle robotic voice, synchronized with its small mouth light. Preserve their appearance, positions and warm lighting. End in an embarrassed pause; no music. No text, logos or watermarks.",
        "duration": 3
      }
    ],
    "aspect_ratio": "16:9"
  }
}
```

<video />

[Download this video](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-pro/text-to-video/v1/01/output.mp4)

### Completed result from the status endpoint

`GET /api/generate/status/{task_id}`

The response below records this verified example. For a new generation, query the task ID returned by your own submission.

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "BRDSCRKN4Q0WH4T6",
    "status": "finished",
    "files": [
      {
        "file_type": "video",
        "file_url": "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-pro/text-to-video/v1/01/output.mp4"
      }
    ],
    "created_time": "2026-09-22T18:43:52",
    "error_message": null,
    "progress": 100
  }
}
```

> [!NOTE]
>
> The API validates prompts up to 2,500 characters. Some multi-shot generation requests have failed when a shot prompt exceeded 512 characters. We recommend keeping each shot prompt within 512 characters; this recommendation does not change the API validation limit.


# Kling O3 Standard Image To Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-o3-std-image-to-video

api-manual/video-series/kwaivgi-kling-video-o3-std-image-to-video.json POST /api/generate/submit
Generate 3–15-second videos with Kling O3 Standard from a starting image and an optional ending frame. Supports single-shot and multi-shot generation. Explicitly set sound; multi-shot requires audio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Parameters

Use model `kwaivgi/kling-video-o3-std/image-to-video`. Put generation parameters inside `input`.

* **`image_urls`**: Required: an array of 1–2 images. The first is the starting frame; the optional second is the ending frame. Accepts public HTTP(S) URLs, image Data URIs or raw Base64 image data.
* **`prompt`**: Required when multi\_shots=false. Describe the scene, action and camera movement. Use nonblank text, up to 2,500 characters after trimming. Omit prompt when multi\_shots=true.
* **`multi_prompt`**: Required when multi\_shots=true; omit in single-shot mode. Supply at least one shot, each with a nonblank prompt of up to 2,500 characters and an integer duration of 1–12 seconds. Shot durations must sum to the top-level duration (3–15 seconds). Extra fields inside a shot are ignored.
* **`duration`**: Required integer from 3 to 15 seconds. In multi-shot mode, this must equal the sum of all shot durations. Credits are calculated using this value.
* **`multi_shots`**: Required: explicitly send false for a single shot or true for multiple shots. Single-shot mode requires prompt. Multi-shot mode requires multi\_prompt and sound=true, with no nonblank top-level prompt. Omitting this field is an error.
* **`sound`**: Required: explicitly send true to generate audio or false for video without audio. Single-shot mode accepts either value; multi-shot mode requires true.
* **`aspect_ratio`**: May be omitted. If provided, use 16:9, 9:16 or 1:1. Image-to-video ignores this value; framing is determined by the input image.

Use standard JSON numbers and booleans. For compatibility, integer strings such as "5" are accepted. Boolean strings true/1/yes/y/on mean true; false/0/no/n/off mean false. These strings are case-insensitive and trimmed. Numeric 1 and 0 are also accepted for boolean fields. Numeric and boolean prompt values are converted to text; 0, false and null are treated as empty. Objects and arrays are not accepted as prompts. Unsupported input fields are rejected. HTTP(S) URLs must have a hostname and no embedded credentials.

## Billing Factors

Billing depends on the model tier, video duration (`duration`), and whether audio is generated (`sound`). Usage is charged per generated second.

Failed generation tasks are refunded.

## Default example: The Miniature Drawbridge

This is the same default example shown on the [model page](https://vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/image-to-video).

```json theme={null}
{
  "model": "kwaivgi/kling-video-o3-std/image-to-video",
  "input": {
    "duration": 3,
    "sound": false,
    "multi_shots": false,
    "prompt": "Fixed-camera miniature scene. Move from the supplied raised-drawbridge start frame to the lowered-drawbridge end frame. The single rigid wooden bridge slowly rotates down around its bottom hinge at the doorway until it reaches the opposite bank; both chains extend naturally. Preserve the towers, moat, tabletop, lighting and framing. No people, new structures or camera movement. No text, logos or watermarks.",
    "image_urls": [
      "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/image-to-video/v1/03/input-start.png",
      "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/image-to-video/v1/03/input-end.png"
    ]
  }
}
```

![Starting frame](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/image-to-video/v1/03/input-start.png)

![Ending frame](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/image-to-video/v1/03/input-end.png)

<video />

[Download this video](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/image-to-video/v1/03/output.mp4)

### Completed result from the status endpoint

`GET /api/generate/status/{task_id}`

The response below records this verified example. For a new generation, query the task ID returned by your own submission.

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "97JVN8DM9N5UJLYO",
    "status": "finished",
    "files": [
      {
        "file_type": "video",
        "file_url": "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/image-to-video/v1/03/output.mp4"
      }
    ],
    "created_time": "2026-09-22T18:58:33",
    "error_message": null,
    "progress": 100
  }
}
```

> [!NOTE]
>
> The API validates prompts up to 2,500 characters. Some multi-shot generation requests have failed when a shot prompt exceeded 512 characters. We recommend keeping each shot prompt within 512 characters; this recommendation does not change the API validation limit.


# Kling O3 Standard Text To Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-o3-std-text-to-video

api-manual/video-series/kwaivgi-kling-video-o3-std-text-to-video.json POST /api/generate/submit
Generate 3–15-second videos from text prompts with Kling O3 Standard. Supports single-shot and multi-shot generation. Explicitly set sound; multi-shot requires audio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Parameters

Use model `kwaivgi/kling-video-o3-std/text-to-video`. Put generation parameters inside `input`.

* **`prompt`**: Required when multi\_shots=false. Describe the scene, action and camera movement. Use nonblank text, up to 2,500 characters after trimming. Omit prompt when multi\_shots=true.
* **`multi_prompt`**: Required when multi\_shots=true; omit in single-shot mode. Supply at least one shot, each with a nonblank prompt of up to 2,500 characters and an integer duration of 1–12 seconds. Shot durations must sum to the top-level duration (3–15 seconds). Extra fields inside a shot are ignored.
* **`duration`**: Required integer from 3 to 15 seconds. In multi-shot mode, this must equal the sum of all shot durations. Credits are calculated using this value.
* **`multi_shots`**: Required: explicitly send false for a single shot or true for multiple shots. Single-shot mode requires prompt. Multi-shot mode requires multi\_prompt and sound=true, with no nonblank top-level prompt. Omitting this field is an error.
* **`sound`**: Required: explicitly send true to generate audio or false for video without audio. Single-shot mode accepts either value; multi-shot mode requires true.
* **`aspect_ratio`**: Optional: 16:9 (landscape), 9:16 (portrait) or 1:1 (square). Sets the video aspect ratio.

Use standard JSON numbers and booleans. For compatibility, integer strings such as "5" are accepted. Boolean strings true/1/yes/y/on mean true; false/0/no/n/off mean false. These strings are case-insensitive and trimmed. Numeric 1 and 0 are also accepted for boolean fields. Numeric and boolean prompt values are converted to text; 0, false and null are treated as empty. Objects and arrays are not accepted as prompts. Unsupported input fields are rejected.

## Billing Factors

Billing depends on the model tier, video duration (`duration`), and whether audio is generated (`sound`). Usage is charged per generated second.

Failed generation tasks are refunded.

## Default example: A Wrench in Weightlessness

This is the same default example shown on the [model page](https://vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/text-to-video).

```json theme={null}
{
  "model": "kwaivgi/kling-video-o3-std/text-to-video",
  "input": {
    "duration": 5,
    "sound": true,
    "multi_shots": true,
    "multi_prompt": [
      {
        "prompt": "Photoreal cinematic science-fiction inside a compact orbital maintenance module. Medium shot: one adult astronaut with short dark hair, wearing a plain off-white flight suit with orange elbow patches, gently catches a single small silver wrench floating in front of her chest. Her body and the wrench drift in zero gravity. Soft blue Earth light through a round window, warm cabin practicals. Quiet ventilation hum and fabric rustle. Controlled natural movement. No text, logos or watermarks.",
        "duration": 2
      },
      {
        "prompt": "Cut to a close view of the same astronaut's hand and orange elbow patch in the same orbital cabin. She places the same silver wrench onto a dark magnetic tool panel. The wrench attaches with one crisp metallic click; she releases it and her hand floats gently away. Maintain the same lighting, costume and wrench. End on the stationary attached wrench. Only quiet ventilation and the synchronized magnetic click; no music or speech. No text, logos or watermarks.",
        "duration": 3
      }
    ],
    "aspect_ratio": "16:9"
  }
}
```

<video />

[Download this video](https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/text-to-video/v1/01/output.mp4)

### Completed result from the status endpoint

`GET /api/generate/status/{task_id}`

The response below records this verified example. For a new generation, query the task ID returned by your own submission.

```json theme={null}
{
  "code": 200,
  "data": {
    "task_id": "57TMAHB6XPG6VHL9",
    "status": "finished",
    "files": [
      {
        "file_type": "video",
        "file_url": "https://cdn.vidgo.ai/apis/models/kwaivgi/kling-video-o3-std/text-to-video/v1/01/output.mp4"
      }
    ],
    "created_time": "2026-09-22T18:37:04",
    "error_message": null,
    "progress": 100
  }
}
```

> [!NOTE]
>
> The API validates prompts up to 2,500 characters. Some multi-shot generation requests have failed when a shot prompt exceeded 512 characters. We recommend keeping each shot prompt within 512 characters; this recommendation does not change the API validation limit.


# Kling 1.6 Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-v1-6-pro-image-to-video

api-manual/video-series/kwaivgi-kling-video-v1-6-pro-image-to-video.json POST /api/generate/submit
Kling 1.6 Pro Image to Video parameters and billing factors.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Public model ID: `kwaivgi/kling-video/v1.6/pro/image-to-video`.

## Input modes

Provide **either** `start_image_url` for first-frame animation **or** `image_urls` for Elements. The endpoint does not accept text-only requests.

* Elements accepts 1–4 reference images and cannot be combined with `start_image_url`, `end_image_url`, or `cfg_scale`.
* Pro supports an optional `end_image_url` when `start_image_url` is present; aspect ratio is supported in both workflows.

The image fields select the workflow.

## Parameters

Place generation parameters inside `input`. Keep `model` and optional `callback_url` at the request root. Unsupported input fields are rejected before credits are deducted.

| Parameter | Requirement | Description |
| - | - | - |
| `prompt` | Required | Required nonblank string, at most 2,500 Unicode characters after trimming. |
| `duration` | Required | Required integer: 5 or 10 seconds. No API default. |
| `aspect_ratio` | Optional | Optional: 1:1, 16:9 or 9:16. No API default. |
| `negative_prompt` | Optional | Optional string, at most 2,500 Unicode characters. |
| `cfg_scale` | Optional | Optional finite number from 0 to 1. Not supported with image\_urls. No API default. |
| `start_image_url` | Conditional | First-frame HTTP(S) URL. Required for the first-frame workflow; mutually exclusive with image\_urls. |
| `end_image_url` | Optional | Optional Pro last-frame HTTP(S) URL. Requires start\_image\_url; not supported with image\_urls. |
| `image_urls` | Conditional | Elements: 1 to 4 HTTP(S) reference URLs. Mutually exclusive with start\_image\_url, end\_image\_url and cfg\_scale. |

Image URLs must be public HTTP(S) URLs with a hostname, without embedded credentials or whitespace. Local files can be uploaded in the playground before submitting their URL. Optional parameters have no documented API defaults.

## Billing Factors

Billing depends on the selected model tier (Standard or Pro) and video duration (`duration`). Usage is charged per generated second. Within a tier, text-to-video, first-frame animation, and Elements use the same rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 1.6 Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-v1-6-pro-text-to-video

api-manual/video-series/kwaivgi-kling-video-v1-6-pro-text-to-video.json POST /api/generate/submit
Kling 1.6 Pro Text to Video parameters and billing factors.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Public model ID: `kwaivgi/kling-video/v1.6/pro/text-to-video`.

## Input modes

This endpoint generates videos from text prompts.

## Parameters

Place generation parameters inside `input`. Keep `model` and optional `callback_url` at the request root. Unsupported input fields are rejected before credits are deducted.

| Parameter | Requirement | Description |
| - | - | - |
| `prompt` | Required | Required nonblank string, at most 2,500 Unicode characters after trimming. |
| `duration` | Required | Required integer: 5 or 10 seconds. No API default. |
| `aspect_ratio` | Optional | Optional: 1:1, 16:9 or 9:16. No API default. |
| `negative_prompt` | Optional | Optional string, at most 2,500 Unicode characters. |
| `cfg_scale` | Optional | Optional finite number from 0 to 1. No API default. |

Optional parameters have no documented API defaults.

## Billing Factors

Billing depends on the selected model tier (Standard or Pro) and video duration (`duration`). Usage is charged per generated second. Within a tier, text-to-video, first-frame animation, and Elements use the same rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 1.6 Standard Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-v1-6-standard-image-to-video

api-manual/video-series/kwaivgi-kling-video-v1-6-standard-image-to-video.json POST /api/generate/submit
Kling 1.6 Standard Image to Video parameters and billing factors.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Public model ID: `kwaivgi/kling-video/v1.6/standard/image-to-video`.

## Input modes

Provide **either** `start_image_url` for first-frame animation **or** `image_urls` for Elements. The endpoint does not accept text-only requests.

* Elements accepts 1–4 reference images and cannot be combined with `start_image_url` or `cfg_scale`.
* Standard first-frame animation does not support `aspect_ratio`; Elements does.

The image fields select the workflow.

## Parameters

Place generation parameters inside `input`. Keep `model` and optional `callback_url` at the request root. Unsupported input fields are rejected before credits are deducted.

| Parameter | Requirement | Description |
| - | - | - |
| `prompt` | Required | Required nonblank string, at most 2,500 Unicode characters after trimming. |
| `duration` | Required | Required integer: 5 or 10 seconds. No API default. |
| `aspect_ratio` | Optional | Optional: 1:1, 16:9 or 9:16. No API default. Not supported in Standard first-frame image-to-video; supported in Elements. |
| `negative_prompt` | Optional | Optional string, at most 2,500 Unicode characters. |
| `cfg_scale` | Optional | Optional finite number from 0 to 1. Not supported with image\_urls. No API default. |
| `start_image_url` | Conditional | First-frame HTTP(S) URL. Required for the first-frame workflow; mutually exclusive with image\_urls. |
| `image_urls` | Conditional | Elements: 1 to 4 HTTP(S) reference URLs. Mutually exclusive with start\_image\_url and cfg\_scale. |

Image URLs must be public HTTP(S) URLs with a hostname, without embedded credentials or whitespace. Local files can be uploaded in the playground before submitting their URL. Optional parameters have no documented API defaults.

## Billing Factors

Billing depends on the selected model tier (Standard or Pro) and video duration (`duration`). Usage is charged per generated second. Within a tier, text-to-video, first-frame animation, and Elements use the same rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 1.6 Standard Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-v1-6-standard-text-to-video

api-manual/video-series/kwaivgi-kling-video-v1-6-standard-text-to-video.json POST /api/generate/submit
Kling 1.6 Standard Text to Video parameters and billing factors.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Public model ID: `kwaivgi/kling-video/v1.6/standard/text-to-video`.

## Input modes

This endpoint generates videos from text prompts.

## Parameters

Place generation parameters inside `input`. Keep `model` and optional `callback_url` at the request root. Unsupported input fields are rejected before credits are deducted.

| Parameter | Requirement | Description |
| - | - | - |
| `prompt` | Required | Required nonblank string, at most 2,500 Unicode characters after trimming. |
| `duration` | Required | Required integer: 5 or 10 seconds. No API default. |
| `aspect_ratio` | Optional | Optional: 1:1, 16:9 or 9:16. No API default. |
| `negative_prompt` | Optional | Optional string, at most 2,500 Unicode characters. |
| `cfg_scale` | Optional | Optional finite number from 0 to 1. No API default. |

Optional parameters have no documented API defaults.

## Billing Factors

Billing depends on the selected model tier (Standard or Pro) and video duration (`duration`). Usage is charged per generated second. Within a tier, text-to-video, first-frame animation, and Elements use the same rate.

Credits are deducted on submission and refunded if the task fails.


# Kling 2.1 Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-v2-1-pro-image-to-video

api-manual/video-series/kwaivgi-kling-video-v2-1-pro-image-to-video.json POST /api/generate/submit
Kling 2.1 Pro image-to-video with a required start frame and an optional end frame.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Kling 2.1 Pro image-to-video with a required start frame and an optional end frame.

## Models

* **kwaivgi/kling-video/v2.1/pro/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Use `start_image_url` to supply the starting frame.
* Optionally supply an ending frame with `end_image_url`.

## Duration Options

* **5, 10 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds** when omitted. Strings, booleans, null and fractional values are rejected.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must be a string and must not be blank after trimming. Non-string values are rejected.

### Start Image URL

* `start_image_url`: Required. Provide the starting frame URL. Use a publicly accessible HTTP(S) URL without credentials or whitespace. Empty URLs and null are rejected.

### End Image URL

* `end_image_url`: Optional. Provide the ending frame URL. Use a publicly accessible HTTP(S) URL without credentials or whitespace. Empty URLs and null are rejected.

### Negative Prompt

* `negative_prompt`: Optional. Describe content to avoid.

## Billing Factors

Billed per video according to the selected model tier (Standard or Pro) and video duration (`duration`).

Credits are deducted on submission and refunded if the task fails.


# Kling 2.1 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/kwaivgi-kling-video-v2-1-standard-image-to-video

api-manual/video-series/kwaivgi-kling-video-v2-1-standard-image-to-video.json POST /api/generate/submit
Kling 2.1 Standard image-to-video with a required start frame.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Kling 2.1 Standard image-to-video with a required start frame.

## Models

* **kwaivgi/kling-video/v2.1/standard/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Use `start_image_url` to supply the starting frame.
* For an optional ending frame, use the [Pro endpoint](/api-manual/video-series/kwaivgi-kling-video-v2-1-pro-image-to-video).

## Duration Options

* **5, 10 seconds** - Set `duration` as an integer. Optional. Default: **5 seconds** when omitted. Strings, booleans, null and fractional values are rejected.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must be a string and must not be blank after trimming. Non-string values are rejected.

### Start Image URL

* `start_image_url`: Required. Provide the starting frame URL. Use a publicly accessible HTTP(S) URL without credentials or whitespace. Empty URLs and null are rejected.

### Negative Prompt

* `negative_prompt`: Optional. Describe content to avoid.

## Billing Factors

Billed per video according to the selected model tier (Standard or Pro) and video duration (`duration`).

Credits are deducted on submission and refunded if the task fails.


# MiniMax H3 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-h3-image-to-video

api-manual/video-series/minimax-h3-image-to-video.json POST /api/generate/submit
MiniMax H3 image to video model with fixed 2K output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **minimax/h3/image-to-video** - MiniMax H3 image to video generation with 2K output

## Features

* **Image To Video**: Generate from one starting frame and an optional ending frame.
* **2K Output**: Output resolution is fixed at `2K`.

## Input modes

* `image_urls` is required.
* All media URLs must be publicly accessible and directly downloadable.

## Parameter Notes

* `prompt` is required and accepts up to 2,000 characters.
* `duration` accepts an integer from `5` through `15` seconds and defaults to `5`.
* `resolution` accepts only `2K`.
* The output aspect ratio follows the first image.
* `image_urls` accepts one starting frame and one optional ending frame.

## Billing

Video generation is billed by output duration.


# MiniMax H3 Max Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-h3-max-image-to-video

api-manual/video-series/minimax-h3-max-image-to-video.json POST /api/generate/submit
The MiniMax H3 Max Image to Video API animates a starting frame and optional ending frame, given as public URLs, into a 5–15 second 480p, 768p or 1080p video.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `enable_safety_checker`: boolean, defaults to `true`. Set `false` to disable it.
* `callback_url`: optional webhook URL at the request root. Query task status if you do not use a callback.

## Usage Guide

Use `minimax/h3-max/image-to-video`. Provide `image_urls` in order: the starting frame first, then an optional ending frame. The output follows the starting image; omit `aspect_ratio`.

Set `model` and optional `callback_url` at the request root. Place generation parameters inside `input`.

## Parameter Details

* `prompt`: required non-empty string, up to 50,000 characters.
* `duration`: integer from 5 through 15 seconds; defaults to 5. Fractional durations are rejected.
* `resolution`: `480p`, `768p` or `1080p`; defaults to `768p`.
* `image_urls`: required array of 1-2 images, in starting-frame and ending-frame order. Use public HTTP(S) download URLs. Base64 is not supported.


# MiniMax H3 Max Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-h3-max-reference-to-video

api-manual/video-series/minimax-h3-max-reference-to-video.json POST /api/generate/submit
Create a video from image, video and audio references.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `enable_safety_checker`: boolean, defaults to `true`. Set `false` to disable it.
* `callback_url`: optional webhook URL at the request root. Query task status if you do not use a callback.

## Usage Guide

Use `minimax/h3-max/reference-to-video`. Provide at least one reference image. Optionally add reference video and audio, with at most 12 assets across all lists. Refer to assets as `Image 1`, `Video 1` and `Audio 1` in list order. Starting and ending frames cannot be combined with reference assets.

Set `model` and optional `callback_url` at the request root. Place generation parameters inside `input`.

## Parameter Details

* `prompt`: required non-empty string, up to 50,000 characters.
* `duration`: integer from 5 through 15 seconds; defaults to 5. Fractional durations are rejected.
* `resolution`: `480p`, `768p` or `1080p`; defaults to `768p`.
* `aspect_ratio`: six fixed ratios (`21:9`, `16:9`, `4:3`, `1:1`, `3:4`, `9:16`) or `adaptive`; defaults to `adaptive`.
* `reference_image_urls`: required array of 1-9 static JPEG, PNG or WebP images, at most 30 MB each. Supported shapes are `1:1`, `4:3`, `3:4`, `16:9`, `9:16`, `5:2` and `2:5`; normalize image orientation before uploading.
* `reference_video_urls`: optional array of up to 3 MP4/MOV clips, at most 50 MB each. Each clip must be 2-15 seconds, with at most 15 seconds combined. Clips must use constant 24 fps, square pixels, zero rotation metadata, and a `16:9` or `9:16` shape. After limiting frames to the requested output duration, the counted frames must be 48, 120, 240 or 360. Other timing patterns are rejected.
* `reference_audio_urls`: optional array of up to 3 MP3/WAV clips, at most 15 MB each. Each clip must be 2-15 seconds and no longer than the output duration, with at most 15 seconds combined. Audio must be mono or stereo at 32 kHz, with a decoded sample count divisible by 800.
* All reference URLs must be public HTTP(S) download links. Total assets across the three lists must not exceed 12.


# MiniMax H3 Max Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-h3-max-text-to-video

api-manual/video-series/minimax-h3-max-text-to-video.json POST /api/generate/submit
The MiniMax H3 Max Text to Video API turns a prompt into a 5–15 second video at 480p, 768p or 1080p, in six aspect ratios from 21:9 to 9:16.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `enable_safety_checker`: boolean, defaults to `true`. Set `false` to disable it.
* `callback_url`: optional webhook URL at the request root. Query task status if you do not use a callback.

## Usage Guide

Use `minimax/h3-max/text-to-video`. Provide a prompt without media. Select the text endpoint for prompt-only generation.

Set `model` and optional `callback_url` at the request root. Place generation parameters inside `input`.

## Parameter Details

* `prompt`: required non-empty string, up to 50,000 characters.
* `duration`: integer from 5 through 15 seconds; defaults to 5. Fractional durations are rejected.
* `resolution`: `480p`, `768p` or `1080p`; defaults to `768p`.
* `aspect_ratio`: `21:9`, `16:9`, `4:3`, `1:1`, `3:4` or `9:16`; defaults to `16:9`.


# MiniMax H3 Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-h3-reference-to-video

api-manual/video-series/minimax-h3-reference-to-video.json POST /api/generate/submit
MiniMax H3 reference to video model with fixed 2K output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **minimax/h3/reference-to-video** - MiniMax H3 reference to video generation with 2K output

## Features

* **Reference To Video**: Guide generation with reference images, videos, and optional audio.
* **2K Output**: Output resolution is fixed at `2K`.

## Input modes

* Provide at least one reference image or reference video. Reference audio cannot be used alone.
* All media URLs must be publicly accessible and directly downloadable.

## Parameter Notes

* `prompt` is required and accepts up to 2,000 characters.
* `duration` accepts an integer from `5` through `15` seconds and defaults to `5`.
* `resolution` accepts only `2K`.
* `aspect_ratio` supports `adaptive`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, and `9:16`.
* Reference input accepts up to 9 images, 3 video clips, and 3 audio clips. Reference audio requires at least one reference image or reference video.

## Billing

Video generation is billed by duration. Each reference-video duration, rounded down separately, is added to output duration; reference images after the first five incur an additional usage charge.


# MiniMax H3 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-h3-text-to-video

api-manual/video-series/minimax-h3-text-to-video.json POST /api/generate/submit
MiniMax H3 text to video model with fixed 2K output

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **minimax/h3/text-to-video** - MiniMax H3 text to video generation with 2K output

## Features

* **Text To Video**: Generate a video directly from a prompt.
* **2K Output**: Output resolution is fixed at `2K`.

## Input modes

* All media URLs must be publicly accessible and directly downloadable.

## Parameter Notes

* `prompt` is required and accepts up to 2,000 characters.
* `duration` accepts an integer from `5` through `15` seconds and defaults to `5`.
* `resolution` accepts only `2K`.
* `aspect_ratio` supports `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, and `9:16`.

## Billing

Video generation is billed by output duration.


# Hailuo 2.3 Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-hailuo-2-3-pro-image-to-video

api-manual/video-series/minimax-hailuo-2-3-pro-image-to-video.json POST /api/generate/submit
Turn a starting image into a 1080p video with Hailuo 2.3 Pro, using text and optional prompt refinement to guide the motion.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn a starting image into a 1080p video with Hailuo 2.3 Pro, using text and optional prompt refinement to guide the motion.

## Models

* **minimax/hailuo-2.3/pro/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Use `start_image_url` to supply the starting frame.
* This model ID fixes the output resolution to `1080p`.

## Duration Options

* **6 seconds** - Set `duration` as an integer. Optional. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. JSON types are validated without converting strings to numbers or booleans; `prompt` must be a string and fractional durations are rejected. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Resolution

* `resolution`: Optional. Accepted values: `1080p`. Default: `1080p`.

### Start Image URL

* `start_image_url`: Required. Provide the starting frame URL. Use a publicly accessible HTTP(S) URL.

### Prompt Optimizer

* `prompt_optimizer`: Optional. Control automatic prompt optimization. Use `true` or `false`. No default is specified; omit this field when unused.

## Billing Factors

Billed per video using the Pro model tier. This endpoint fixes output resolution to `1080p` and duration to `6` seconds. Prompt optimization (`prompt_optimizer`) does not change the charge.

Credits are deducted on submission and refunded if the task fails.


# Hailuo 2.3 Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-hailuo-2-3-pro-text-to-video

api-manual/video-series/minimax-hailuo-2-3-pro-text-to-video.json POST /api/generate/submit
Generate short video scenes with Hailuo 2.3 Pro, with 1080p output and optional prompt refinement for written scene concepts.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate short video scenes with Hailuo 2.3 Pro, with 1080p output and optional prompt refinement for written scene concepts.

## Models

* **minimax/hailuo-2.3/pro/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* This model ID fixes the output resolution to `1080p`.

## Duration Options

* **6 seconds** - Set `duration` as an integer. Optional. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. JSON types are validated without converting strings to numbers or booleans; `prompt` must be a string and fractional durations are rejected. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Resolution

* `resolution`: Optional. Accepted values: `1080p`. Default: `1080p`.

### Prompt Optimizer

* `prompt_optimizer`: Optional. Control automatic prompt optimization. Use `true` or `false`. No default is specified; omit this field when unused.

## Billing Factors

Billed per video using the Pro model tier. This endpoint fixes output resolution to `1080p` and duration to `6` seconds. Prompt optimization (`prompt_optimizer`) does not change the charge.

Credits are deducted on submission and refunded if the task fails.


# Hailuo 2.3 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-hailuo-2-3-standard-image-to-video

api-manual/video-series/minimax-hailuo-2-3-standard-image-to-video.json POST /api/generate/submit
Turn a starting image into a 768p video with Hailuo 2.3 Standard, using text and optional prompt refinement to guide the motion.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn a starting image into a 768p video with Hailuo 2.3 Standard, using text and optional prompt refinement to guide the motion.

## Models

* **minimax/hailuo-2.3/standard/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Use `start_image_url` to supply the starting frame.
* This model ID fixes the output resolution to `768p`.

## Duration Options

* **6, 10 seconds** - Set `duration` as an integer. Optional. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. JSON types are validated without converting strings to numbers or booleans; `prompt` must be a string and fractional durations are rejected. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Resolution

* `resolution`: Optional. Accepted values: `768p`. Default: `768p`.

### Start Image URL

* `start_image_url`: Required. Provide the starting frame URL. Use a publicly accessible HTTP(S) URL.

### Prompt Optimizer

* `prompt_optimizer`: Optional. Control automatic prompt optimization. Use `true` or `false`. No default is specified; omit this field when unused.

## Billing Factors

Billing depends on video duration (`duration`) using the Standard model tier. This endpoint fixes output resolution to `768p`. Prompt optimization (`prompt_optimizer`) does not change the charge.

Credits are deducted on submission and refunded if the task fails.


# Hailuo 2.3 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-hailuo-2-3-standard-text-to-video

api-manual/video-series/minimax-hailuo-2-3-standard-text-to-video.json POST /api/generate/submit
Generate short video scenes with Hailuo 2.3 Standard, with 768p output and optional prompt refinement for written scene concepts.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate short video scenes with Hailuo 2.3 Standard, with 768p output and optional prompt refinement for written scene concepts.

## Models

* **minimax/hailuo-2.3/standard/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* This model ID fixes the output resolution to `768p`.

## Duration Options

* **6, 10 seconds** - Set `duration` as an integer. Optional. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. JSON types are validated without converting strings to numbers or booleans; `prompt` must be a string and fractional durations are rejected. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.
* The value must not be blank when supplied.

### Resolution

* `resolution`: Optional. Accepted values: `768p`. Default: `768p`.

### Prompt Optimizer

* `prompt_optimizer`: Optional. Control automatic prompt optimization. Use `true` or `false`. No default is specified; omit this field when unused.

## Billing Factors

Billing depends on video duration (`duration`) using the Standard model tier. This endpoint fixes output resolution to `768p`. Prompt optimization (`prompt_optimizer`) does not change the charge.

Credits are deducted on submission and refunded if the task fails.


# Wan 2.6
Source: https://docs.vidgo.ai/api-manual/video-series/wan-2-6

api-manual/video-series/wan-2-6.json POST /api/generate/submit
Multi-shot 1080p video generation with Alibaba Wan 2.6

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Wan 2.6 is Alibaba's latest video generation model family. It supports text-to-video, image-to-video, and video-to-video workflows with stable character identity, multi-shot composition, and 1080p output.

## Models

* **wan2.6-text-to-video** - Generate videos from text prompts
* **wan2.6-image-to-video** - Animate a reference image with a prompt
* **wan2.6-video-to-video** - Edit or restyle an input video using a prompt

## Features

* **Duration**: 5, 10, or 15 seconds (video-to-video up to 10 seconds)
* **Resolution**: 720p or 1080p
* **Multi-Shots**: Toggle multi-shot composition for cinematic transitions
* **Prompt Length**: Up to 5,000 characters

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [alibaba/wan-2.6/text-to-video](/api-manual/video-series/alibaba-wan-2-6-text-to-video)
* [alibaba/wan-2.6/image-to-video](/api-manual/video-series/alibaba-wan-2-6-image-to-video)
* [alibaba/wan-2.6/reference-to-video](/api-manual/video-series/alibaba-wan-2-6-reference-to-video)


# Wan 2.2 Image-to-Video Fast
Source: https://docs.vidgo.ai/api-manual/video-series/wan2.2-image-to-video-fast

api-manual/video-series/wan2.2-image-to-video-fast.json POST /api/generate/submit
Fast image-to-video generation with Alibaba Wan 2.2 Fast

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

`wan2.2-image-to-video-fast` animates one reference image, and optionally uses a second image as the last frame for stronger transition control.

## Models

* **wan2.2-image-to-video-fast** - Animate a reference image into a short video

## Key Parameters

* **Image URLs**: 1 required, up to 2 supported
* **Resolution**: `480p` or `720p`
* **Seed**: optional integer for reproducible outputs


# Wan 2.2 Text-to-Video Fast
Source: https://docs.vidgo.ai/api-manual/video-series/wan2.2-text-to-video-fast

api-manual/video-series/wan2.2-text-to-video-fast.json POST /api/generate/submit
Fast text-to-video generation with Alibaba Wan 2.2 Fast

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

`wan2.2-text-to-video-fast` is optimized for fast text-to-video generation with supported portrait and landscape output options.

## Models

* **wan2.2-text-to-video-fast** - Generate short videos from text prompts

## Key Parameters

* **Aspect Ratio**: `16:9`, `9:16`
* **Resolution**: `480p` or `720p`
* **Seed**: optional integer for reproducible outputs


# Wan 2.5 Image-to-Video
Source: https://docs.vidgo.ai/api-manual/video-series/wan2.5-image-to-video

api-manual/video-series/wan2.5-image-to-video.json POST /api/generate/submit
Image-to-video generation with Alibaba Wan 2.5

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

`wan2.5-image-to-video` is a higher-quality single-image animation model with broader resolution support than the Wan 2.2 fast line.

## Models

* **wan2.5-image-to-video** - Animate a single image into a video clip

## Key Parameters

* **Image URLs**: exactly 1 required
* **Resolution**: `480p`, `720p`, or `1080p`
* **Duration**: `5` or `10` seconds
* **Optional Controls**: `audio`, `negative_prompt`, `seed`

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [alibaba/wan-2.5/image-to-video](/api-manual/video-series/alibaba-wan-2-5-image-to-video)


# Wan 2.5 Text-to-Video
Source: https://docs.vidgo.ai/api-manual/video-series/wan2.5-text-to-video

api-manual/video-series/wan2.5-text-to-video.json POST /api/generate/submit
Text-to-video generation with Alibaba Wan 2.5

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

`wan2.5-text-to-video` is the text-only branch of the Wan 2.5 video family and supports multiple size presets with 5s and 10s generation lengths.

## Models

* **wan2.5-text-to-video** - Generate a video clip from a text prompt

## Key Parameters

* **Aspect Ratio**: `832*480`, `480*832`, `1280*720`, `720*1280`, `1920*1080`, `1080*1920`
* **Duration**: `5` or `10` seconds
* **Optional Controls**: `audio` string, `negative_prompt`, `seed`

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [alibaba/wan-2.5/text-to-video](/api-manual/video-series/alibaba-wan-2-5-text-to-video)

## Notes

* If `audio` is provided, it must be a string. Do not send a boolean value.


# Grok Imagine Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/xai-grok-imagine-video-image-to-video

api-manual/video-series/xai-grok-imagine-video-image-to-video.json POST /api/generate/submit
Animate a still image with Grok Imagine, using text to shape the scene in six- or ten-second video clips.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate a still image with Grok Imagine, using text to shape the scene in six- or ten-second video clips.

## Models

* **xai/grok-imagine-video/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Accepts exactly 1 image in `image_urls`.

## Duration Options

* **6, 10 seconds** - Set `duration` as an integer. Optional; defaults to `6`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below. Supply publicly accessible HTTP(S) URLs for reference images.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.

### Mode

* `mode`: Optional. Accepted values: `fun`, `normal`, `spicy`.


# Grok Imagine Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/xai-grok-imagine-video-text-to-video

api-manual/video-series/xai-grok-imagine-video-text-to-video.json POST /api/generate/submit
Create six- or ten-second video scenes with Grok Imagine, using text to guide the action in square, portrait, or landscape formats.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create six- or ten-second video scenes with Grok Imagine, using text to guide the action in square, portrait, or landscape formats.

## Models

* **xai/grok-imagine-video/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.

## Duration Options

* **6, 10 seconds** - Set `duration` as an integer. Optional; defaults to `6`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input fields listed below.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-5,000 characters**.

### Aspect Ratio

* `aspect_ratio`: Optional. Accepted values: `1:1`, `2:3`, `3:2`, `16:9`, `9:16`.

### Mode

* `mode`: Optional. Accepted values: `fun`, `normal`, `spicy`.


# Grok Imagine Video 1.5 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/xai-grok-imagine-video-v1-5-image-to-video

api-manual/video-series/xai-grok-imagine-video-v1-5-image-to-video.json POST /api/generate/submit
Animate one first-frame image with Grok Imagine Video 1.5. Generate 1–15 second videos at 480p, 720p, or 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Grok Imagine Video 1.5 Image to Video

Animate a single image using a prompt that describes subject motion, camera movement, and sound. The image supplies the first frame and determines the video's aspect ratio.

## Request

Set `model` to `xai/grok-imagine-video-v1.5/image-to-video`. Place the following parameters inside `input`. Set optional `callback_url` at the request root.

### prompt

Required string. Describe the scene and desired motion in **1–4,096 characters**, including at least one non-whitespace character.

### image\_urls

Required array containing **exactly one** publicly accessible HTTP(S) image URL. Use a URL with a hostname and credential-free access.

### resolution

Optional string: `480p`, `720p`, or `1080p`. Default: `720p`.

### duration

Optional integer from **1 to 15** seconds. Default: **6**.


# Hunyuan 3D v3.1 Pro Image to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/hunyuan-3d-v3-1-pro-image-to-3d

api-manual/3d-series/hunyuan-3d-v3-1-pro-image-to-3d.json POST /api/generate/submit
Reconstruct 3D objects with Hunyuan 3D v3.1 Pro using a primary image and additional views, with support for PBR materials.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Reconstruct 3D objects with Hunyuan 3D v3.1 Pro using a primary image and additional views, with support for PBR materials.

## Models

* **hunyuan-3d/v3.1/pro/image-to-3d** - Generate a 3D model from an image.

## Features

* Generate a 3D model from an image.
* Accepts exactly 1 image in `image_urls`.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Generate Type

* `generate_type`: Optional. Choose `Normal` for a textured model or `Geometry` for a geometry-only white model.

### Face Count

* `face_count`: Optional. Set the target face count. Integer from `40000` to `1500000`. Default: `500000`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input parameters listed below. Provide publicly accessible HTTP(S) URLs for image inputs.

### Prompt

* `prompt`: Optional. Describe the desired 3D object. Minimum length: **1 character**.
* Use a nonblank string when providing this parameter.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.

### PBR

* `enable_pbr`: Optional. Set `true` to generate PBR materials in textured mode. Geometry-only generation takes precedence over this setting.

### Back Image URL

* `back_image_url`: Optional. Provide the back view as an image URL. Use a publicly accessible HTTP(S) URL.

### Left Image URL

* `left_image_url`: Optional. Provide the left view as an image URL. Use a publicly accessible HTTP(S) URL.

### Right Image URL

* `right_image_url`: Optional. Provide the right view as an image URL. Use a publicly accessible HTTP(S) URL.

### Top Image URL

* `top_image_url`: Optional. Provide the top view as an image URL. Use a publicly accessible HTTP(S) URL.

### Bottom Image URL

* `bottom_image_url`: Optional. Provide the bottom view as an image URL. Use a publicly accessible HTTP(S) URL.

### Left Front Image URL

* `left_front_image_url`: Optional. Provide the left front view as an image URL. Use a publicly accessible HTTP(S) URL.

### Right Front Image URL

* `right_front_image_url`: Optional. Provide the right front view as an image URL. Use a publicly accessible HTTP(S) URL.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Additional artifacts may use `file_type: "other"`. Retain returned textures, material files, and file metadata.


# Hunyuan 3D v3.1 Pro Text to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/hunyuan-3d-v3-1-pro-text-to-3d

api-manual/3d-series/hunyuan-3d-v3-1-pro-text-to-3d.json POST /api/generate/submit
Generate 3D objects from text with Hunyuan 3D v3.1 Pro, with control over mesh complexity and geometry-only output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate 3D objects from text with Hunyuan 3D v3.1 Pro, with control over mesh complexity and geometry-only output.

## Models

* **hunyuan-3d/v3.1/pro/text-to-3d** - Generate a 3D model from a text prompt.

## Features

* Generate a 3D model from a text prompt.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Generate Type

* `generate_type`: Optional. Choose `Normal` for a textured model or `Geometry` for a geometry-only white model.

### Face Count

* `face_count`: Optional. Set the target face count. Integer from `40000` to `1500000`. Default: `500000`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input parameters listed below.

### Prompt

* `prompt`: Required. Describe the desired 3D object. Length: **1-1,024 characters**.
* Use a nonblank string when providing this parameter.

### PBR

* `enable_pbr`: Optional. Set `true` to generate PBR materials in textured mode. Geometry-only generation takes precedence over this setting.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Additional artifacts may use `file_type: "other"`. Retain returned textures, material files, and file metadata.


# Hunyuan 3D v3.1 Rapid Image to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/hunyuan-3d-v3-1-rapid-image-to-3d

api-manual/3d-series/hunyuan-3d-v3-1-rapid-image-to-3d.json POST /api/generate/submit
Create a 3D model from a reference image with Hunyuan 3D v3.1 Rapid, with optional text guidance and geometry generation.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create a 3D model from a reference image with Hunyuan 3D v3.1 Rapid, with optional text guidance and geometry generation.

## Models

* **hunyuan-3d/v3.1/rapid/image-to-3d** - Generate a 3D model from an image.

## Features

* Generate a 3D model from an image.
* Accepts exactly 1 image in `image_urls`.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Enable Geometry

* `enable_geometry`: Optional. Set `true` for a geometry-only white model or `false` for textured generation.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input parameters listed below. Provide publicly accessible HTTP(S) URLs for image inputs.

### Prompt

* `prompt`: Optional. Describe the desired 3D object. Minimum length: **1 character**.
* Use a nonblank string when providing this parameter.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.

### PBR

* `enable_pbr`: Optional. Set `true` to generate PBR materials in textured mode. Geometry-only generation takes precedence over this setting.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Additional artifacts may use `file_type: "other"`. Retain returned textures, material files, and file metadata.


# Hunyuan 3D v3.1 Rapid Text to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/hunyuan-3d-v3-1-rapid-text-to-3d

api-manual/3d-series/hunyuan-3d-v3-1-rapid-text-to-3d.json POST /api/generate/submit
Turn written object concepts into 3D models with Hunyuan 3D v3.1 Rapid, with an option to generate geometry for further modeling.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn written object concepts into 3D models with Hunyuan 3D v3.1 Rapid, with an option to generate geometry for further modeling.

## Models

* **hunyuan-3d/v3.1/rapid/text-to-3d** - Generate a 3D model from a text prompt.

## Features

* Generate a 3D model from a text prompt.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Enable Geometry

* `enable_geometry`: Optional. Set `true` for a geometry-only white model or `false` for textured generation.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input parameters listed below.

### Prompt

* `prompt`: Required. Describe the desired 3D object. Length: **1-200 characters**.
* Use a nonblank string when providing this parameter.

### PBR

* `enable_pbr`: Optional. Set `true` to generate PBR materials in textured mode. Geometry-only generation takes precedence over this setting.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Additional artifacts may use `file_type: "other"`. Retain returned textures, material files, and file metadata.


# Meshy 6 Image to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/meshy-v6-image-to-3d

api-manual/3d-series/meshy-v6-image-to-3d.json POST /api/generate/submit
Turn a reference image into a textured 3D asset with Meshy 6, with remeshing, material controls, and optional rigging and animation.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn a reference image into a textured 3D asset with Meshy 6, with remeshing, material controls, and optional rigging and animation.

## Models

* **meshy/v6/image-to-3d** - Generate a 3D model from an image.

## Features

* Generate a 3D model from an image.
* Accepts exactly 1 image in `image_urls`.
* Retrieve model assets and available previews or supporting files from the task result.
* Optionally add textures, rigging, and animation.

## Output Options

### Target Polycount

* `target_polycount`: Optional. Set the target polygon count. Integer. Default: `30000`.

### Topology

* `topology`: Optional. Select the mesh topology. Accepted values: `quad`, `triangle`. Default: `triangle`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Place the parameters listed below inside `input`. Use publicly accessible HTTP(S) URLs for images and texture references.

* `enable_animation: true` requires `enable_rigging: true`.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.

### Remeshing

* `should_remesh`: Optional. Control remeshing. Use `true` or `false`. Default: `true`.

### Symmetry Mode

* `symmetry_mode`: Optional. Select how symmetry is handled. Accepted values: `off`, `auto`, `on`. Default: `auto`.

### Textures

* `should_texture`: Optional. Control texture generation. Use `true` or `false`. Default: `true`.

### PBR

* `enable_pbr`: Optional. Control physically based rendering (PBR) materials. Use `true` or `false`.

### Pose Mode

* `pose_mode`: Optional. Select a pose, or send an empty string for no pose selection. Accepted values: `a-pose`, `t-pose`, `""`. Default: `""`.

### Texture Prompt

* `texture_prompt`: Optional. Describe the desired texture.

### Texture Image URL

* `texture_image_url`: Optional. Provide a texture reference URL. Use a publicly accessible HTTP(S) URL.

### Rigging

* `enable_rigging`: Optional. Request rigging for the generated model. Use `true` or `false`.

### Rigging Height

* `rigging_height_meters`: Optional. Set the rigging height in meters. Number. Default: `1.7`.

### Animation

* `enable_animation`: Optional. Request animation; rigging must also be enabled. Use `true` or `false`.

### Animation Action ID

* `animation_action_id`: Optional. Select the animation action. Integer. Default: `1001`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`. Default: `true`.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Read supporting files with `file_type: "other"` alongside the returned textures, animation assets, and file metadata.


# Meshy 6 Multi Image to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/meshy-v6-multi-image-to-3d

api-manual/3d-series/meshy-v6-multi-image-to-3d.json POST /api/generate/submit
Build a 3D asset from up to four reference images with Meshy 6, with texture, topology, and optional animation controls.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Build a 3D asset from up to four reference images with Meshy 6, with texture, topology, and optional animation controls.

## Models

* **meshy/v6/multi-image-to-3d** - Generate a 3D model from multiple image references.

## Features

* Generate a 3D model from multiple image references.
* Accepts 1-4 images in `image_urls`.
* Retrieve model assets and available previews or supporting files from the task result.
* Optionally add textures, rigging, and animation.

## Output Options

### Target Polycount

* `target_polycount`: Optional. Set the target polygon count. Integer. Default: `30000`.

### Topology

* `topology`: Optional. Select the mesh topology. Accepted values: `quad`, `triangle`. Default: `triangle`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Place the parameters listed below inside `input`. Use publicly accessible HTTP(S) URLs for images and texture references.

* `enable_animation: true` requires `enable_rigging: true`.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply 1-4 items. Each item must be an HTTP(S) URL.

### Remeshing

* `should_remesh`: Optional. Control remeshing. Use `true` or `false`. Default: `true`.

### Symmetry Mode

* `symmetry_mode`: Optional. Select how symmetry is handled. Accepted values: `off`, `auto`, `on`. Default: `auto`.

### Textures

* `should_texture`: Optional. Control texture generation. Use `true` or `false`. Default: `true`.

### PBR

* `enable_pbr`: Optional. Control physically based rendering (PBR) materials. Use `true` or `false`.

### Pose Mode

* `pose_mode`: Optional. Select a pose, or send an empty string for no pose selection. Accepted values: `a-pose`, `t-pose`, `""`. Default: `""`.

### Texture Prompt

* `texture_prompt`: Optional. Describe the desired texture.

### Texture Image URL

* `texture_image_url`: Optional. Provide a texture reference URL. Use a publicly accessible HTTP(S) URL.

### Rigging

* `enable_rigging`: Optional. Request rigging for the generated model. Use `true` or `false`.

### Rigging Height

* `rigging_height_meters`: Optional. Set the rigging height in meters. Number. Default: `1.7`.

### Animation

* `enable_animation`: Optional. Request animation; rigging must also be enabled. Use `true` or `false`.

### Animation Action ID

* `animation_action_id`: Optional. Select the animation action. Integer. Default: `1001`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`. Default: `true`.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Read supporting files with `file_type: "other"` alongside the returned textures, animation assets, and file metadata.


# Meshy 6 Text to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/meshy-v6-text-to-3d

api-manual/3d-series/meshy-v6-text-to-3d.json POST /api/generate/submit
Build 3D assets from text with Meshy 6, with mesh topology, material, and texture controls plus optional rigging and animation.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Build 3D assets from text with Meshy 6, with mesh topology, material, and texture controls plus optional rigging and animation.

## Models

* **meshy/v6/text-to-3d** - Generate a 3D model from a text prompt.

## Features

* Generate a 3D model from a text prompt.
* Retrieve model assets and available previews or supporting files from the task result.
* Optionally add textures, rigging, and animation.

## Output Options

### Mode

* `mode`: Optional. Accepted values: `preview`, `full`. Default: `full`.

### Target Polycount

* `target_polycount`: Optional. Set the target polygon count. Integer. Default: `30000`.

### Topology

* `topology`: Optional. Select the mesh topology. Accepted values: `quad`, `triangle`. Default: `triangle`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Place the parameters listed below inside `input`. Use publicly accessible HTTP(S) URLs for images and texture references.

* `enable_animation: true` requires `enable_rigging: true`.

### Prompt

* `prompt`: Required. Describe the desired 3D object. Length: **1-600 characters**.
* Write a description of the object, its shape, and its materials.

### Seed

* `seed`: Optional. Set the generation seed. Integer.

### Remeshing

* `should_remesh`: Optional. Control remeshing. Use `true` or `false`. Default: `true`.

### Symmetry Mode

* `symmetry_mode`: Optional. Select how symmetry is handled. Accepted values: `off`, `auto`, `on`. Default: `auto`.

### PBR

* `enable_pbr`: Optional. Control physically based rendering (PBR) materials. Use `true` or `false`.

### Pose Mode

* `pose_mode`: Optional. Select a pose, or send an empty string for no pose selection. Accepted values: `a-pose`, `t-pose`, `""`. Default: `""`.

### Enable Prompt Expansion

* `enable_prompt_expansion`: Optional. Control automatic prompt expansion. Use `true` or `false`.

### Texture Prompt

* `texture_prompt`: Optional. Describe the desired texture.

### Texture Image URL

* `texture_image_url`: Optional. Provide a texture reference URL. Use a publicly accessible HTTP(S) URL.

### Rigging

* `enable_rigging`: Optional. Request rigging for the generated model. Use `true` or `false`.

### Rigging Height

* `rigging_height_meters`: Optional. Set the rigging height in meters. Number. Default: `1.7`.

### Animation

* `enable_animation`: Optional. Request animation; rigging must also be enabled. Use `true` or `false`.

### Animation Action ID

* `animation_action_id`: Optional. Select the animation action. Integer. Default: `1001`.

### Safety Checker

* `enable_safety_checker`: Optional. Control safety checking. Use `true` or `false`. Default: `true`.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Read supporting files with `file_type: "other"` alongside the returned textures, animation assets, and file metadata.


# Meshy 7.1 Image to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/meshy-v7-1-image-to-3d

api-manual/3d-series/meshy-v7-1-image-to-3d.json POST /api/generate/submit
The Meshy 7.1 Image to 3D API turns one photo or concept image into a textured 3D model with up to 4K geometry, Smart Topology, PBR maps and rigging.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Meshy 7.1 Image to 3D reconstructs a 3D model from a single image, such as a product photo or a character concept. It is textured by default, supports up to `4k` geometry detail, and can produce clean low-poly meshes, PBR maps and rigged characters.

Meshy 7.1 is Meshy's 3D generation model, released with a focus on richer geometric detail. Vidgo exposes it as three model IDs: [Text to 3D](/api-manual/3d-series/meshy-v7-1-text-to-3d), [Image to 3D](/api-manual/3d-series/meshy-v7-1-image-to-3d) and [Multi-Image to 3D](/api-manual/3d-series/meshy-v7-1-multi-image-to-3d).

## Models

* **meshy/v7.1/image-to-3d** — Image-to-3D with Meshy 7.1 from exactly one image.

## Features

* Image-to-3D from one photo or concept image.
* Textured by default; set `should_texture: false` for geometry only.
* Geometry detail up to `4k` with `model_type: standard`, or low-poly meshes with `smart-topology`.
* Triangle or quad topology and a target polygon count.
* PBR maps, texture guidance by text or image.
* Optional humanoid rigging and animation presets.

## Required parameters

* `image_urls` (string\[]): Exactly one image, as a public HTTP(S) URL or a `data:image/...` URI. Must be an array. Sharp, well-lit JPG or PNG images with one centered subject work best.

## Optional parameters

* `should_texture` (boolean): Texture the model. Default `true`; `false` returns geometry only.
* `model_type` (string): `standard` or `smart-topology`. Default `standard`.
* `topology` (string): `triangle` or `quad`. Default `triangle`.
* `target_polycount` (integer): `100`–`300000`, default `30000`. With `model_type: smart-topology` the range is `100`–`15000`, default `4000`.
* `should_remesh` (boolean): Remesh the output. Default `true`; `false` keeps the raw triangle mesh.
* `symmetry_mode` (string): `off`, `auto` or `on`. Default `auto`.
* `pose_mode` (string): `""` (no pose), `a-pose` or `t-pose`. Default `""`.
* `geometry_resolution` (string): `standard`, `2k` or `4k`. Only with `model_type: standard`.
* `enable_pbr` (boolean): Add metallic, roughness and normal maps to the base color texture. Default `false`.
* `texture_prompt` (string): Text guidance for the texture, up to 600 characters.
* `texture_image_url` (string): Image that guides the texture (public HTTP(S) URL or `data:image/...` URI).
* `enable_rigging` (boolean): Auto-rig the model as a humanoid character. Default `false`.
* `rigging_height_meters` (number): Character height used for rigging, greater than 0. Default `1.7`.
* `enable_animation` (boolean): Apply an animation preset. Requires `enable_rigging: true`. Default `false`.
* `animation_action_id` (integer): Animation preset ID, `0`–`696`. Default `92`.
* `enable_safety_checker` (boolean): Upstream safety checker. Default `true`.

## Output

Read every item in the `files` array returned by [Query Task Status](/api-manual/task-management/status) or the callback. Model assets use `file_type: "3d"` (typically GLB, plus FBX, OBJ, USDZ or STL when available), previews and thumbnails use `file_type: "image"`, and textures, rigged or animated assets and other supporting files may use `file_type: "other"`. Not every format is returned for every task, so do not assume a fixed file list.

## Examples

* **Textured model from a product photo**: one image with `geometry_resolution: 4k` and PBR maps.
* **Low-poly game asset**: `model_type: smart-topology`, quad topology and a 4,000 polygon target.
* **Rigged character**: a character concept in `a-pose` with `enable_rigging` and a `callback_url`.

## Notes

* `image_urls` takes exactly one image. For 2–4 views of the same object, use [Meshy 7.1 Multi-Image to 3D](/api-manual/3d-series/meshy-v7-1-multi-image-to-3d); to start from text, use [Meshy 7.1 Text to 3D](/api-manual/3d-series/meshy-v7-1-text-to-3d).
* `geometry_resolution` with `model_type: smart-topology` returns 400, as does a `target_polycount` above 15,000 with `smart-topology`. `geometry_resolution` sets mesh detail, not texture image size.
* `enable_animation: true` without `enable_rigging: true` returns 400. Rigging works best on clearly limbed humanoid characters in `a-pose` or `t-pose`; check rigs and meshes before using them in production.
* Fields not listed above, such as `prompt`, `mode` or `seed`, return 400. Use `texture_prompt` for text guidance.
* Credits are reserved at submission: a base charge per model, plus extra for texturing (`should_texture: true`, the default), `enable_rigging` and `enable_animation`. Polycount, topology, `geometry_resolution` and image count do not change the price. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Meshy 7.1 Multi-Image to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/meshy-v7-1-multi-image-to-3d

api-manual/3d-series/meshy-v7-1-multi-image-to-3d.json POST /api/generate/submit
The Meshy 7.1 Multi-Image to 3D API reconstructs a 3D model from 1–4 views of the same object, with up to 2K geometry, topology control, PBR maps and rigging.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Meshy 7.1 Multi-Image to 3D reconstructs one 3D model from up to four views of the same object, such as front, side and back photos. Extra angles give the model more information about hidden sides than a single image can. Use it for product scans, props and character turnarounds.

Meshy 7.1 is Meshy's 3D generation model, released with a focus on richer geometric detail. Vidgo exposes it as three model IDs: [Text to 3D](/api-manual/3d-series/meshy-v7-1-text-to-3d), [Image to 3D](/api-manual/3d-series/meshy-v7-1-image-to-3d) and [Multi-Image to 3D](/api-manual/3d-series/meshy-v7-1-multi-image-to-3d).

## Models

* **meshy/v7.1/multi-image-to-3d** — Multi-view image-to-3D with Meshy 7.1 from 1–4 images.

## Features

* Multi-view image-to-3D from 1–4 images of the same object.
* Textured by default; set `should_texture: false` for geometry only.
* Geometry detail up to `2k`.
* Triangle or quad topology and a target polygon count up to 300,000.
* PBR maps, texture guidance by text or image.
* Optional humanoid rigging and animation presets.

## Required parameters

* `image_urls` (string\[]): 1–4 views of the same object, as public HTTP(S) URLs or `data:image/...` URIs. Use different angles (front, side, back, three-quarter) with consistent lighting.

## Optional parameters

* `should_texture` (boolean): Texture the model. Default `true`; `false` returns geometry only.
* `topology` (string): `triangle` or `quad`. Default `triangle`.
* `target_polycount` (integer): `100`–`300000`. Default `30000`.
* `should_remesh` (boolean): Remesh the output. Default `true`; `false` keeps the raw triangle mesh.
* `symmetry_mode` (string): `off`, `auto` or `on`. Default `auto`.
* `pose_mode` (string): `""` (no pose), `a-pose` or `t-pose`. Default `""`.
* `geometry_resolution` (string): `standard` or `2k`.
* `enable_pbr` (boolean): Add metallic, roughness and normal maps to the base color texture. Default `false`. Requires `should_texture: true`.
* `texture_prompt` (string): Text guidance for the texture, up to 600 characters. Requires `should_texture: true`.
* `texture_image_url` (string): Image that guides the texture (public HTTP(S) URL or `data:image/...` URI). Requires `should_texture: true`.
* `enable_rigging` (boolean): Auto-rig the model as a humanoid character. Default `false`.
* `rigging_height_meters` (number): Character height used for rigging, greater than 0. Default `1.7`.
* `enable_animation` (boolean): Apply an animation preset. Requires `enable_rigging: true`. Default `false`.
* `animation_action_id` (integer): Animation preset ID, `0`–`696`. Default `92`.
* `enable_safety_checker` (boolean): Upstream safety checker. Default `true`.

## Output

Read every item in the `files` array returned by [Query Task Status](/api-manual/task-management/status) or the callback. Model assets use `file_type: "3d"` (typically GLB, plus FBX, OBJ, USDZ or STL when available), previews and thumbnails use `file_type: "image"`, and textures, rigged or animated assets and other supporting files may use `file_type: "other"`. Not every format is returned for every task, so do not assume a fixed file list.

## Examples

* **Textured model from three views**: front, side and back photos of a chair with `geometry_resolution: 2k` and PBR maps.
* **Geometry only**: two views of a vase with `should_texture: false`, quad topology and a `callback_url`.

## Notes

* Multi-image does not support `4k` geometry or `model_type`; use [Meshy 7.1 Image to 3D](/api-manual/3d-series/meshy-v7-1-image-to-3d) for `4k` or Smart Topology from a single image.
* With `should_texture: false`, sending `enable_pbr: true`, `texture_prompt` or `texture_image_url` returns 400.
* `enable_animation: true` without `enable_rigging: true` returns 400. Rigging works best on clearly limbed humanoid characters in `a-pose` or `t-pose`; check rigs and meshes before using them in production.
* Fields not listed above, such as `prompt`, `mode`, `seed` or `model_type`, return 400.
* Credits are reserved at submission: a base charge per model, plus extra for texturing (`should_texture: true`, the default), `enable_rigging` and `enable_animation`. Polycount, topology, `geometry_resolution` and image count do not change the price. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Meshy 7.1 Text to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/meshy-v7-1-text-to-3d

api-manual/3d-series/meshy-v7-1-text-to-3d.json POST /api/generate/submit
The Meshy 7.1 Text to 3D API creates a 3D model from a prompt, in preview or textured mode, with up to 4K geometry, quad topology, PBR maps and rigging.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Meshy 7.1 Text to 3D turns a written description into a 3D model. Choose a fast geometry-only `preview` or a textured `full` model, control topology and polygon count for game engines, and optionally rig and animate humanoid characters.

Meshy 7.1 is Meshy's 3D generation model, released with a focus on richer geometric detail. Vidgo exposes it as three model IDs: [Text to 3D](/api-manual/3d-series/meshy-v7-1-text-to-3d), [Image to 3D](/api-manual/3d-series/meshy-v7-1-image-to-3d) and [Multi-Image to 3D](/api-manual/3d-series/meshy-v7-1-multi-image-to-3d).

## Models

* **meshy/v7.1/text-to-3d** — Text-to-3D with Meshy 7.1, preview or textured output.

## Features

* Text-to-3D from a prompt of up to 600 characters.
* `preview` (geometry only) or `full` (textured) mode.
* Geometry detail up to `4k` with `model_type: standard`, or cleaner low-poly meshes with `smart-topology`.
* Triangle or quad topology and a target polygon count.
* PBR maps, texture guidance by text or image.
* Optional humanoid rigging and animation presets.

## Required parameters

* `prompt` (string): The object to create, including shape, materials, colors and style. 1–600 characters, not blank.

## Optional parameters

* `mode` (string): `preview` (geometry only) or `full` (textured). Default `full`.
* `model_type` (string): `standard` or `smart-topology`. Default `standard`.
* `seed` (integer): Seed for repeatable results.
* `enable_prompt_expansion` (boolean): Let the model expand a short prompt. Default `false`.
* `topology` (string): `triangle` or `quad`. Default `triangle`.
* `target_polycount` (integer): `100`–`300000`, default `30000`. With `model_type: smart-topology` the range is `100`–`15000`, default `4000`.
* `should_remesh` (boolean): Remesh the output. Default `true`; `false` keeps the raw triangle mesh.
* `symmetry_mode` (string): `off`, `auto` or `on`. Default `auto`.
* `pose_mode` (string): `""` (no pose), `a-pose` or `t-pose`. Default `""`.
* `geometry_resolution` (string): `standard`, `2k` or `4k`. Only with `model_type: standard`.
* `enable_pbr` (boolean): Add metallic, roughness and normal maps to the base color texture. Default `false`.
* `texture_prompt` (string): Text guidance for the texture, up to 600 characters.
* `texture_image_url` (string): Image that guides the texture (public HTTP(S) URL or `data:image/...` URI).
* `enable_rigging` (boolean): Auto-rig the model as a humanoid character. Default `false`.
* `rigging_height_meters` (number): Character height used for rigging, greater than 0. Default `1.7`.
* `enable_animation` (boolean): Apply an animation preset. Requires `enable_rigging: true`. Default `false`.
* `animation_action_id` (integer): Animation preset ID, `0`–`696`. Default `92`.
* `enable_safety_checker` (boolean): Upstream safety checker. Default `true`.

## Output

Read every item in the `files` array returned by [Query Task Status](/api-manual/task-management/status) or the callback. Model assets use `file_type: "3d"` (typically GLB, plus FBX, OBJ, USDZ or STL when available), previews and thumbnails use `file_type: "image"`, and textures, rigged or animated assets and other supporting files may use `file_type: "other"`. Not every format is returned for every task, so do not assume a fixed file list.

## Examples

* **Textured prop**: `mode: full` with quad topology, a 30,000 polygon target and PBR maps.
* **Geometry-only preview**: `mode: preview` with `geometry_resolution: 2k` and a fixed `seed`, useful for checking shapes before texturing.
* **Rigged and animated character**: a full-body character in `t-pose` with `enable_rigging` and `enable_animation`, plus a `callback_url`.

## Notes

* This ID accepts text only. For a photo or concept image, use [Meshy 7.1 Image to 3D](/api-manual/3d-series/meshy-v7-1-image-to-3d); for several views of one object, use [Meshy 7.1 Multi-Image to 3D](/api-manual/3d-series/meshy-v7-1-multi-image-to-3d).
* `geometry_resolution` with `model_type: smart-topology` returns 400, as does a `target_polycount` above 15,000 with `smart-topology`. `geometry_resolution` sets mesh detail, not texture image size.
* `enable_animation: true` without `enable_rigging: true` returns 400. Rigging works best on clearly limbed humanoid characters in `a-pose` or `t-pose`; check rigs and meshes before using them in production.
* Fields not listed above, such as `image_urls` or `should_texture`, return 400. `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission: a base charge per model, plus extra for texturing (`mode: full`, the default), `enable_rigging` and `enable_animation`. Polycount, topology, `geometry_resolution` and image count do not change the price. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# 3D Models
Source: https://docs.vidgo.ai/api-manual/3d-series/overview

All 3D generation models available on Vidgo API, with their model IDs.

Every 3D model is called through `POST /api/generate/submit`. Send a model ID in `model`, the model's parameters in `input`, and an optional `callback_url`. The request returns a `task_id` right away and the task runs asynchronously.

## Workflow

1. Submit a task to `POST /api/generate/submit` with a model ID from the table below.
2. Poll [Get Task Status](/api-manual/task-management/status) with the returned `task_id`, or receive the result at `callback_url` (see [Webhook Callbacks](/api-manual/task-management/webhooks)).
3. Download the files in `files` when the status is `finished`.

3D tasks can return several files, such as model assets, previews, and textures. Process every item in the `files` array.

## Models

| Model | Provider | Model IDs |
| - | - | - |
| [Meshy 6 Text to 3D](/api-manual/3d-series/meshy-v6-text-to-3d) | Meshy | `meshy/v6/text-to-3d` |
| [Meshy 6 Image to 3D](/api-manual/3d-series/meshy-v6-image-to-3d) | Meshy | `meshy/v6/image-to-3d` |
| [Meshy 6 Multi Image to 3D](/api-manual/3d-series/meshy-v6-multi-image-to-3d) | Meshy | `meshy/v6/multi-image-to-3d` |
| [Meshy 7.1 Text to 3D](/api-manual/3d-series/meshy-v7-1-text-to-3d) | Meshy | `meshy/v7.1/text-to-3d` |
| [Meshy 7.1 Image to 3D](/api-manual/3d-series/meshy-v7-1-image-to-3d) | Meshy | `meshy/v7.1/image-to-3d` |
| [Meshy 7.1 Multi-Image to 3D](/api-manual/3d-series/meshy-v7-1-multi-image-to-3d) | Meshy | `meshy/v7.1/multi-image-to-3d` |
| [Hunyuan 3D v3.1 Pro Text to 3D](/api-manual/3d-series/hunyuan-3d-v3-1-pro-text-to-3d) | Hunyuan3D | `hunyuan-3d/v3.1/pro/text-to-3d` |
| [Hunyuan 3D v3.1 Pro Image to 3D](/api-manual/3d-series/hunyuan-3d-v3-1-pro-image-to-3d) | Hunyuan3D | `hunyuan-3d/v3.1/pro/image-to-3d` |
| [Hunyuan 3D v3.1 Rapid Text to 3D](/api-manual/3d-series/hunyuan-3d-v3-1-rapid-text-to-3d) | Hunyuan3D | `hunyuan-3d/v3.1/rapid/text-to-3d` |
| [Hunyuan 3D v3.1 Rapid Image to 3D](/api-manual/3d-series/hunyuan-3d-v3-1-rapid-image-to-3d) | Hunyuan3D | `hunyuan-3d/v3.1/rapid/image-to-3d` |
| [Tripo3D H3.1 Text to 3D](/api-manual/3d-series/tripo3d-h3-1-text-to-3d) | Tripo3D | `tripo3d/h3.1/text-to-3d` |
| [Tripo3D H3.1 Image to 3D](/api-manual/3d-series/tripo3d-h3-1-image-to-3d) | Tripo3D | `tripo3d/h3.1/image-to-3d` |
| [Tripo3D H3.1 Multiview to 3D](/api-manual/3d-series/tripo3d-h3-1-multiview-to-3d) | Tripo3D | `tripo3d/h3.1/multiview-to-3d` |
| [Tripo3D P1 Text to 3D](/api-manual/3d-series/tripo3d-p1-text-to-3d) | Tripo3D | `tripo3d/p1/text-to-3d` |
| [Tripo3D P1 Image to 3D](/api-manual/3d-series/tripo3d-p1-image-to-3d) | Tripo3D | `tripo3d/p1/image-to-3d` |

## Notes

* Each model page lists the parameters, limits, and modes for its model IDs. Parameters are not shared across models.
* Generated files are valid for 24 hours after creation. Download and store them promptly.
* URL inputs must be publicly accessible HTTP(S) URLs. To use local files, upload them first with [File Management](/api-manual/file-series/upload-url).


# Tripo3D H3.1 Image to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/tripo3d-h3-1-image-to-3d

api-manual/3d-series/tripo3d-h3-1-image-to-3d.json POST /api/generate/submit
Convert a reference image into a 3D model with Tripo3D H3.1, with controls for geometry detail, textures, and mesh topology.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Convert a reference image into a 3D model with Tripo3D H3.1, with controls for geometry detail, textures, and mesh topology.

## Models

* **tripo3d/h3.1/image-to-3d** - Generate a 3D model from an image.

## Features

* Generate a 3D model from an image.
* Accepts exactly 1 image in `image_urls`.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Face Limit

* `face_limit`: Optional. Set the face-count limit. Integer from `1000` to `2000000`.

### Geometry Quality

* `geometry_quality`: Optional. Accepted values: `standard`, `detailed`. Default: `standard`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input parameters listed below for this generation workflow. Use publicly accessible HTTP(S) image URLs with a hostname.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.

### Texture

* `texture`: Optional. Control texture generation. Use `true` or `false`. Default: `true`.

### PBR

* `pbr`: Optional. Generate physically based rendering (PBR) materials when `texture` is `true`. Use `true` or `false`.

### Model Seed

* `model_seed`: Optional. Set the model-generation seed. Integer.

### Texture Seed

* `texture_seed`: Optional. Set the texture-generation seed. Integer.

### Texture Quality

* `texture_quality`: Optional. Accepted values: `standard`, `detailed`. Default: `standard`.

### Auto Size

* `auto_size`: Optional. Control automatic model sizing. Use `true` or `false`.

### Quad

* `quad`: Optional. Request quad topology. Use `true` or `false`. Default: `false`.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Additional artifacts may use `file_type: "other"`. Retain accompanying assets and available `label`, `format`, `content_type`, `file_name`, and `file_size` metadata.


# Tripo3D H3.1 Multiview to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/tripo3d-h3-1-multiview-to-3d

api-manual/3d-series/tripo3d-h3-1-multiview-to-3d.json POST /api/generate/submit
Reconstruct a 3D asset from multiple views with Tripo3D H3.1, with adjustable geometry, textures, and PBR materials.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Reconstruct a 3D asset from multiple views with Tripo3D H3.1, with adjustable geometry, textures, and PBR materials.

## Models

* **tripo3d/h3.1/multiview-to-3d** - Generate a 3D model from multiple views.

## Features

* Generate a 3D model from multiple views.
* Accepts 2-4 images in `image_urls`.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Face Limit

* `face_limit`: Optional. Set the face-count limit. Integer from `1000` to `2000000`.

### Geometry Quality

* `geometry_quality`: Optional. Accepted values: `standard`, `detailed`. Default: `standard`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input parameters listed below for this generation workflow. Use publicly accessible HTTP(S) image URLs with a hostname.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs showing the same object from different angles. Supply 2-4 items. Each item must be an HTTP(S) URL.

### Texture

* `texture`: Optional. Control texture generation. Use `true` or `false`. Default: `true`.

### PBR

* `pbr`: Optional. Generate physically based rendering (PBR) materials when `texture` is `true`. Use `true` or `false`.

### Model Seed

* `model_seed`: Optional. Set the model-generation seed. Integer.

### Texture Seed

* `texture_seed`: Optional. Set the texture-generation seed. Integer.

### Texture Quality

* `texture_quality`: Optional. Accepted values: `standard`, `detailed`. Default: `standard`.

### Auto Size

* `auto_size`: Optional. Control automatic model sizing. Use `true` or `false`.

### Quad

* `quad`: Optional. Request quad topology. Use `true` or `false`. Default: `false`.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Additional artifacts may use `file_type: "other"`. Retain accompanying assets and available `label`, `format`, `content_type`, `file_name`, and `file_size` metadata.


# Tripo3D H3.1 Text to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/tripo3d-h3-1-text-to-3d

api-manual/3d-series/tripo3d-h3-1-text-to-3d.json POST /api/generate/submit
Create textured 3D objects from text with Tripo3D H3.1, with control over mesh detail, surface quality, and PBR materials.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create textured 3D objects from text with Tripo3D H3.1, with control over mesh detail, surface quality, and PBR materials.

## Models

* **tripo3d/h3.1/text-to-3d** - Generate a 3D model from a text prompt.

## Features

* Generate a 3D model from a text prompt.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Face Limit

* `face_limit`: Optional. Set the face-count limit. Integer from `1000` to `2000000`.

### Geometry Quality

* `geometry_quality`: Optional. Accepted values: `standard`, `detailed`. Default: `standard`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the input parameters listed below for this generation workflow.

### Prompt

* `prompt`: Required. Describe the desired 3D object. Length: **1-1,024 characters**.

### Negative Prompt

* `negative_prompt`: Optional. Describe content to avoid. Maximum length: **1,024 characters**.

### Texture

* `texture`: Optional. Control texture generation. Use `true` or `false`. Default: `true`.

### PBR

* `pbr`: Optional. Generate physically based rendering (PBR) materials when `texture` is `true`. Use `true` or `false`.

### Model Seed

* `model_seed`: Optional. Set the model-generation seed. Integer.

### Image Seed

* `image_seed`: Optional. Set the image-generation seed. Integer.

### Texture Seed

* `texture_seed`: Optional. Set the texture-generation seed. Integer.

### Texture Quality

* `texture_quality`: Optional. Accepted values: `standard`, `detailed`. Default: `standard`.

### Auto Size

* `auto_size`: Optional. Control automatic model sizing. Use `true` or `false`.

### Quad

* `quad`: Optional. Request quad topology. Use `true` or `false`. Default: `false`.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Additional artifacts may use `file_type: "other"`. Retain accompanying assets and available `label`, `format`, `content_type`, `file_name`, and `file_size` metadata.


# Tripo3D P1 Image to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/tripo3d-p1-image-to-3d

api-manual/3d-series/tripo3d-p1-image-to-3d.json POST /api/generate/submit
Turn a reference image into a 3D object with Tripo3D P1, with optional textures and control over polygon count.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn a reference image into a 3D object with Tripo3D P1, with optional textures and control over polygon count.

## Models

* **tripo3d/p1/image-to-3d** - Generate a 3D model from an image.

## Features

* Generate a 3D model from an image.
* Accepts exactly 1 image in `image_urls`.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Face Limit

* `face_limit`: Optional. Set the face-count limit. Integer from `48` to `20000`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the `input` fields listed below for this workflow. Provide a publicly accessible HTTP(S) image URL.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.

### Texture

* `texture`: Optional. Control texture generation. Use `true` or `false`. Default: `true`.

### Model Seed

* `model_seed`: Optional. Set the model-generation seed. Integer.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Supporting files use `file_type: "other"`. Preserve each returned file and its `label`, `format`, `content_type`, `file_name`, and `file_size` metadata when present.
* Use the `model_glb` file for 3D preview and the `thumbnail` file as its cover image when these labels are present.


# Tripo3D P1 Text to 3D
Source: https://docs.vidgo.ai/api-manual/3d-series/tripo3d-p1-text-to-3d

api-manual/3d-series/tripo3d-p1-text-to-3d.json POST /api/generate/submit
Create 3D objects from text with Tripo3D P1, with optional textures and a compact mesh budget for props and scene assets.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create 3D objects from text with Tripo3D P1, with optional textures and a compact mesh budget for props and scene assets.

## Models

* **tripo3d/p1/text-to-3d** - Generate a 3D model from a text prompt.

## Features

* Generate a 3D model from a text prompt.
* Retrieve model assets and available previews or supporting files from the task result.

## Output Options

### Face Limit

* `face_limit`: Optional. Set the face-count limit. Integer from `48` to `20000`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the `input` fields listed below for this workflow.

### Prompt

* `prompt`: Required. Describe the desired 3D object. Length: **1-1,024 characters**.
* Provide a non-empty string describing the object.

### Texture

* `texture`: Optional. Control texture generation. Use `true` or `false`. Default: `true`.

### Model Seed

* `model_seed`: Optional. Set the model-generation seed. Integer.

## Output Files

* Read every item in the returned `files` array.
* Model assets use `file_type: "3d"`; previews use `file_type: "image"`.
* Supporting files use `file_type: "other"`. Preserve each returned file and its `label`, `format`, `content_type`, `file_name`, and `file_size` metadata when present.
* Use the `model_glb` file for 3D preview and the `thumbnail` file as its cover image when these labels are present.


# Chat Completions
Source: https://docs.vidgo.ai/api-manual/chat-series/chat-completions

api-manual/chat-series/chat-completions.json POST /v1/chat/completions
OpenAI-compatible Chat Completions API

Use a model that supports the Chat Completions protocol. Set `stream: true` for an SSE response.

```bash theme={null}
curl https://api.vidgo.ai/v1/chat/completions \
  -H "Authorization: Bearer $VIDGO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"kimi-k3","messages":[{"role":"user","content":"Hello"}]}'
```

## Additional supported models

* `openai/gpt-6-astra`
* `anthropic/claude-fable-5.1`
* `xai/grok-4.6`
* `google/gemini-3.7-flash`
* `google/gemini-3.5-flash`
* `anthropic/claude-opus-4.8`
* `openai/gpt-5.5`
* `openai/gpt-5.4`
* `openai/gpt-5.2`
* `anthropic/claude-sonnet-4.6`
* `anthropic/claude-opus-4.6`
* `anthropic/claude-opus-5.5`
* `anthropic/claude-sonnet-5.5`
* `anthropic/claude-haiku-5.5`
* `anthropic/claude-opus-4.7`
* `anthropic/claude-opus-4.5`
* `anthropic/claude-sonnet-4.5`
* `anthropic/claude-haiku-4.5`
* `openai/gpt-6.1-sol`
* `openai/gpt-6.1-terra`
* `openai/gpt-6-sol`
* `openai/gpt-6-luna`
* `google/gemini-3.8-flash`
* `google/gemini-4-argon`
* `google/gemini-3.1-pro-preview`
* `google/gemini-3-flash-preview`
* `deepseek/deepseek-v4-pro`
* `deepseek/deepseek-v4-flash`
* `deepseek/deepseek-v4.1-flash`
* `x-ai/grok-4.7`

Send protocol fields at the root, without a generation input wrapper. messages must be a non-empty array.


# Claude Messages
Source: https://docs.vidgo.ai/api-manual/chat-series/claude-messages

api-manual/chat-series/claude-messages.json POST /v1/messages
Anthropic-compatible Messages API

Use a model that supports the Messages protocol. `Authorization: Bearer` and `x-api-key` authentication are both accepted.

```bash theme={null}
curl https://api.vidgo.ai/v1/messages \
  -H "x-api-key: $VIDGO_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "Content-Type: application/json" \
  -d '{"model":"claude-sonnet-5","max_tokens":512,"messages":[{"role":"user","content":"Hello"}]}'
```

## Additional supported models

* `anthropic/claude-fable-5.1`
* `anthropic/claude-opus-4.8`
* `anthropic/claude-sonnet-4.6`
* `anthropic/claude-opus-4.6`
* `anthropic/claude-opus-5.5`
* `anthropic/claude-sonnet-5.5`
* `anthropic/claude-haiku-5.5`
* `anthropic/claude-opus-4.7`
* `anthropic/claude-opus-4.5`
* `anthropic/claude-sonnet-4.5`
* `anthropic/claude-haiku-4.5`
* `deepseek/deepseek-v4.1-flash`

Send protocol fields at the root, without a generation input wrapper. messages must be a non-empty array. max\_tokens must be a positive integer.


# Gemini Native Format
Source: https://docs.vidgo.ai/api-manual/chat-series/gemini-native-format

api-manual/chat-series/gemini-native-format.json POST /v1beta/models/{model}:generateContent
Gemini-compatible generateContent and streaming endpoints

Use a model that supports the Gemini native protocol. Unsupported model IDs return a Gemini-formatted error response.

For streaming, call `POST /v1beta/models/{model}:streamGenerateContent` and consume the SSE response.

## Additional supported models

* `google/gemini-3.7-flash`
* `google/gemini-3.5-flash`
* `google/gemini-3.8-flash`
* `google/gemini-4-argon`
* `google/gemini-3.1-pro-preview`
* `google/gemini-3-flash-preview`

Use the complete public ID in the URL, including the provider slash. For example: `/v1beta/models/google/gemini-3.7-flash:generateContent`. Send a non-empty contents array. Streaming uses `:streamGenerateContent?alt=sse`.


# LLM Models
Source: https://docs.vidgo.ai/api-manual/chat-series/overview

All large language models available on Vidgo API, with the API formats each model supports.

All Chat Series endpoints use `https://api.vidgo.ai` and a Vidgo API key. Send the key as `Authorization: Bearer VIDGO_API_KEY`. Claude clients may also use `x-api-key`; Gemini clients may use `x-goog-api-key`.

## Model protocol support

| Model | Responses | Chat Completions | Claude Messages | Gemini native |
| - | :-: | :-: | :-: | :-: |
| `gpt-5-6-luna` | Yes | — | — | — |
| `gpt-5-6-terra` | Yes | — | — | — |
| `gpt-5-6-sol` | Yes | — | — | — |
| `kimi-k3` | — | Yes | — | — |
| `claude-sonnet-5` | — | Yes | Yes | — |
| `claude-opus-5` | — | Yes | Yes | — |

The legacy IDs above keep their existing protocol support. Gemini workflows are listed below.

Both normal JSON responses and SSE streaming are supported.

## Additional model protocol support

| Model | Chat Completions | Responses | Claude Messages | Gemini native |
| - | :-: | :-: | :-: | :-: |
| `openai/gpt-6-astra` | Yes | Yes | — | — |
| `anthropic/claude-fable-5.1` | Yes | — | Yes | — |
| `xai/grok-4.6` | Yes | Yes | — | — |
| `google/gemini-3.7-flash` | Yes | — | — | Yes |
| `google/gemini-3.5-flash` | Yes | — | — | Yes |
| `anthropic/claude-opus-4.8` | Yes | — | Yes | — |
| `openai/gpt-5.5` | Yes | Yes | — | — |
| `openai/gpt-5.4` | Yes | Yes | — | — |
| `openai/gpt-5.2` | Yes | Yes | — | — |
| `anthropic/claude-sonnet-4.6` | Yes | — | Yes | — |
| `anthropic/claude-opus-4.6` | Yes | — | Yes | — |
| `anthropic/claude-opus-5.5` | Yes | — | Yes | — |
| `anthropic/claude-sonnet-5.5` | Yes | — | Yes | — |
| `anthropic/claude-haiku-5.5` | Yes | — | Yes | — |
| `anthropic/claude-opus-4.7` | Yes | — | Yes | — |
| `anthropic/claude-opus-4.5` | Yes | — | Yes | — |
| `anthropic/claude-sonnet-4.5` | Yes | — | Yes | — |
| `anthropic/claude-haiku-4.5` | Yes | — | Yes | — |
| `openai/gpt-6.1-sol` | Yes | Yes | — | — |
| `openai/gpt-6.1-terra` | Yes | Yes | — | — |
| `openai/gpt-6-sol` | Yes | Yes | — | — |
| `openai/gpt-6-luna` | Yes | Yes | — | — |
| `google/gemini-3.8-flash` | Yes | — | — | Yes |
| `google/gemini-4-argon` | Yes | — | — | Yes |
| `google/gemini-3.1-pro-preview` | Yes | — | — | Yes |
| `google/gemini-3-flash-preview` | Yes | — | — | Yes |
| `deepseek/deepseek-v4-pro` | Yes | — | — | — |
| `deepseek/deepseek-v4-flash` | Yes | — | — | — |
| `deepseek/deepseek-v4.1-flash` | Yes | Yes | Yes | — |
| `x-ai/grok-4.7` | Yes | — | — | — |

Response bodies and SSE events retain the upstream protocol format. A returned model name may be the upstream name rather than the Vidgo public ID.


# Responses
Source: https://docs.vidgo.ai/api-manual/chat-series/responses

api-manual/chat-series/responses.json POST /v1/responses
OpenAI-compatible Responses API

Use a model that supports the Responses protocol. Set `stream: true` for SSE.

```bash theme={null}
curl https://api.vidgo.ai/v1/responses \
  -H "Authorization: Bearer $VIDGO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5-6-luna","input":"Explain Decimal arithmetic briefly."}'
```

## Additional supported models

* `openai/gpt-6-astra`
* `xai/grok-4.6`
* `openai/gpt-5.5`
* `openai/gpt-5.4`
* `openai/gpt-5.2`
* `openai/gpt-6.1-sol`
* `openai/gpt-6.1-terra`
* `openai/gpt-6-sol`
* `openai/gpt-6-luna`
* `deepseek/deepseek-v4.1-flash`

Send protocol fields at the root, without a generation input wrapper. input must not be empty.


# Add Instrumental
Source: https://docs.vidgo.ai/api-manual/music-series/add-instrumental

api-manual/music-series/add-instrumental.json POST /api/generate/submit
Generate musical accompaniment for uploaded audio files

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint generates musical accompaniment for uploaded audio files
* Typically used with vocal stems or melodies to create backing tracks
* Perfect for adding instrumental layers to existing recordings without needing a producer

## Parameter Details

* Required parameters:

  * `upload_url`: URL of the audio file to process

  * `title`: Name for the generated track

  * `tags`: Desired musical styles (e.g., "relaxing, piano, soothing")

  * `negative_tags`: Styles to exclude (e.g., "heavy metal, fast drums")

* Model selection:

  * **V5**: Superior musical expressiveness with faster generation

  * **V4\_5PLUS**: Enhanced tonal richness (default)

## Usage Notes

* Use descriptive tags to guide the style of the instrumental
* Combine positive and negative tags for precise control over the output
* The callback mechanism sends notifications at three stages: text generation, first track completion, and all tracks completed

## Optional parameters

* `model` (string): AI model version. Options: `V4_5PLUS` (default), `V5`.

* `vocal_gender` (string): Vocal gender preference. Use `m` for male, `f` for female. Note: This parameter can only increase the probability but cannot guarantee the specified gender.

* `style_weight` (number): Strength of adherence to style. Range 0-1, up to 2 decimals.

* `weirdness_constraint` (number): Controls creative deviation. Range 0-1, up to 2 decimals.

* `audio_weight` (number): Balance weight for audio features. Range 0-1, up to 2 decimals.


# Add Vocals
Source: https://docs.vidgo.ai/api-manual/music-series/add-vocals

api-manual/music-series/add-vocals.json POST /api/generate/submit
Layer AI-generated vocals on top of an existing instrumental

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint adds AI-generated vocals to an existing instrumental track
* Upload an instrumental file and provide lyrics or a description for the vocals
* Perfect for transforming instrumental tracks into complete songs

## Parameter Details

* Required parameters:

  * `upload_url`: URL of the instrumental audio file

  * `prompt`: Lyrics or description for the vocals

  * `title`: Name for the generated track

  * `style`: Music style to guide vocal delivery

  * `negative_tags`: Styles to exclude from generation

* Model selection:

  * **V5**: Superior musical expressiveness with faster generation

  * **V4\_5PLUS**: Enhanced tonal richness (default)

## Usage Notes

* The `prompt` parameter serves as both lyrics and creative direction for the vocals
* Use `style` to specify the vocal delivery style (e.g., "Pop Ballad", "Rock", "R\&B")
* Combine with `vocal_gender` for more control over the vocal output

## Optional parameters

* `model` (string): AI model version. Options: `V4_5PLUS` (default), `V5`.

* `vocal_gender` (string): Vocal gender preference. Use `m` for male, `f` for female. Note: This parameter can only increase the probability but cannot guarantee the specified gender.

* `style_weight` (number): Strength of adherence to style. Range 0-1, up to 2 decimals.

* `weirdness_constraint` (number): Controls creative deviation. Range 0-1, up to 2 decimals.

* `audio_weight` (number): Balance weight for audio features. Range 0-1, up to 2 decimals.


# Audio Models
Source: https://docs.vidgo.ai/api-manual/music-series/audio-overview

All text-to-speech (TTS) models available on Vidgo API, with their model IDs.

Every text-to-speech model is called through `POST /api/generate/submit`. Send a model ID in `model`, the model's parameters in `input`, and an optional `callback_url`. The request returns a `task_id` right away and the task runs asynchronously.

## Workflow

1. Submit a task to `POST /api/generate/submit` with a model ID from the table below.
2. Poll [Get Task Status](/api-manual/task-management/status) with the returned `task_id`, or receive the result at `callback_url` (see [Webhook Callbacks](/api-manual/task-management/webhooks)).
3. Download the files in `files` when the status is `finished`.

## Models

| Model | Provider | Model IDs |
| - | - | - |
| [ElevenLabs V3 TTS](/api-manual/music-series/elevenlabs-tts-eleven-v3) | ElevenLabs | `elevenlabs/tts/eleven-v3` |
| [ElevenLabs TTS Turbo v2.5](/api-manual/music-series/elevenlabs-tts-turbo-v2-5) | ElevenLabs | `elevenlabs/tts/turbo-v2.5` |
| [ElevenLabs V4 TTS](/api-manual/music-series/elevenlabs-tts-eleven-v4) | ElevenLabs | `elevenlabs/tts/eleven-v4` |
| [ElevenLabs V4 Turbo TTS](/api-manual/music-series/elevenlabs-tts-eleven-v4-turbo) | ElevenLabs | `elevenlabs/tts/eleven-v4-turbo` |
| [Gemini 3.1 Flash TTS](/api-manual/music-series/google-gemini-3-1-flash-text-to-speech) | Google | `google/gemini-3.1-flash/text-to-speech` |
| [Gemini 3.8 Flash TTS](/api-manual/music-series/google-gemini-3-8-flash-text-to-speech) | Google | `google/gemini-3.8-flash/text-to-speech` |
| [Gemini 3.8 Flash Lite TTS](/api-manual/music-series/google-gemini-3-8-flash-lite-text-to-speech) | Google | `google/gemini-3.8-flash-lite/text-to-speech` |
| [xAI TTS 1](/api-manual/music-series/xai-tts-v1) | xAI | `xai/tts/v1` |

## Notes

* Each model page lists the parameters, limits, and modes for its model IDs. Parameters are not shared across models.
* Generated files are valid for 24 hours after creation. Download and store them promptly.
* URL inputs must be publicly accessible HTTP(S) URLs. To use local files, upload them first with [File Management](/api-manual/file-series/upload-url).


# Boost Music Style
Source: https://docs.vidgo.ai/api-manual/music-series/boost-music-style

api-manual/music-series/boost-music-style.json POST /api/generate/submit
Generate enhanced music style descriptions using AI

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint generates enhanced music style descriptions based on your input
* Provide concise keywords describing genre, mood, or musical characteristics
* The AI will expand and enrich your style description for better music generation results

## Parameter Details

* `content` (required): A brief description of the desired music style
  * Use comma-separated keywords for best results
  * Include genre (Pop, Jazz, Rock), mood (Mysterious, Upbeat), or instruments (Piano, Guitar)

## Usage Notes

* Use the generated style descriptions with the Generate Music endpoint for better results
* Keep input descriptions concise and clear for optimal enhancement


# Convert to WAV
Source: https://docs.vidgo.ai/api-manual/music-series/convert-to-wav

api-manual/music-series/convert-to-wav.json POST /api/generate/submit
Convert generated music tracks to high-quality WAV format

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint converts generated music tracks to high-quality WAV format
* WAV files provide lossless audio quality for professional use
* The download URL for the WAV file is delivered via the callback

## Parameter Details

* `task_id` (required): Task ID from a completed music generation (Generate Music or Extend Music)
* `audio_id` (required): Specific audio track identifier from the callback data

## Usage Notes

* Each audio track can only be converted to WAV once
* The callback will include a download URL for the high-quality WAV file
* WAV files are significantly larger than compressed formats - plan storage accordingly


# Create Music Video
Source: https://docs.vidgo.ai/api-manual/music-series/create-music-video

api-manual/music-series/create-music-video.json POST /api/generate/submit
Generate visualized music videos from audio tracks

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint generates visualized music videos from audio tracks
* Creates MP4 videos with audio visualization and optional branding
* Useful for social media content, promotional materials, or music sharing

## Parameter Details

* `task_id` (required): Task ID from a completed music generation (Generate Music or Extend Music)
* `audio_id` (required): Specific audio track identifier from the callback data

## Usage Notes

* Each audio track can only have one music video generated
* If a video already exists, the API returns error code 409
* The callback includes a download URL for the generated MP4 video
* Video generation may take longer than audio processing tasks

## Optional parameters

* `author` (string): Artist or creator name displayed on the video cover. Maximum 50 characters.
* `domain_name` (string): Website or brand watermark displayed at the bottom of the video. Maximum 50 characters.


# ElevenLabs Music
Source: https://docs.vidgo.ai/api-manual/music-series/elevenlabs-music

api-manual/music-series/elevenlabs-music.json POST /api/generate/submit
Compose songs or instrumentals from a text brief or a section-by-section plan, with control over musical style, lyrics, and arrangement.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Compose songs or instrumentals from a text brief or a section-by-section plan, with control over musical style, lyrics, and arrangement.

## Models

* **elevenlabs/music** - Generate music from text or a structured composition plan.

## Features

* Generate music from text or a structured composition plan.
* Plan musical sections with their own styles, lyrics, and durations.
* Choose an output encoding with `output_format`.

## Duration Options

* **Text input** - Optional `duration` is a number from **3 to 600 seconds**. Use it only with `text`.
* **Composition plan** - Each section accepts **3-120 seconds**.

## Output Options

### Output Format

* `output_format`: Optional. String. Default: `mp3_44100_128`.

> [accordion] Supported output format values
>
> * `mp3_22050_32`
> * `mp3_44100_32`
> * `mp3_44100_64`
> * `mp3_44100_96`
> * `mp3_44100_128`
> * `mp3_44100_192`
> * `pcm_8000`
> * `pcm_16000`
> * `pcm_22050`
> * `pcm_24000`
> * `pcm_44100`
> * `pcm_48000`
> * `ulaw_8000`
> * `alaw_8000`
> * `opus_48000_32`
> * `opus_48000_64`
> * `opus_48000_96`
> * `opus_48000_128`
> * `opus_48000_192`

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Choose one creative path: a text brief or a structured composition plan.

* Provide exactly one of `text` or `composition_plan`.
* `duration` and `is_instrumental` apply only to `text`. Use `respect_sections_durations` only with `composition_plan`.

### Text

* `text`: Required for the text path. Describe the music to generate. Length: **1-4,100 characters**.
* Include a music description with the desired genre, mood, and instruments.

### Composition Plan

* `composition_plan`: Required for the composition-plan path. Describe global styles and an ordered list of musical sections. Supply an object with the fields below.

* `composition_plan.positive_global_styles`: Required. Supply an array. Each item must be a string.

* `composition_plan.negative_global_styles`: Required. Supply an array. Each item must be a string.

* `composition_plan.sections`: Required. Supply at least 1 item. Each item is an object.

* `composition_plan.sections[].section_name`: Required. Length: **1-100 characters**.

* `composition_plan.sections[].positive_local_styles`: Required. Supply an array. Each item must be a string.

* `composition_plan.sections[].negative_local_styles`: Required. Supply an array. Each item must be a string.

* `composition_plan.sections[].duration`: Required. Duration in seconds. Number from `3` to `120`.

* `composition_plan.sections[].lines`: Required. Supply an array. Each item must be a string. Each string is limited to **200 characters**.

* Section durations are measured in seconds.

### Instrumental Mode

* `is_instrumental`: Optional. Request instrumental music with `true`. Default: `false` for the text path.
* Use only with `text`.

### Section Timing

* `respect_sections_durations`: Optional. Use only with a composition plan to control adherence to section durations. Use `true` or `false`. Default: `true` for the composition-plan path.

## Output Files

* Read every item in the returned `files` array.
* Generated audio uses `file_type: "audio"`.
* Timestamp and alignment files may use `file_type: "other"`. Preserve optional file metadata when present.

## Examples

Each example pairs a request with its completed task response and generated audio. Submit the request to `POST https://api.vidgo.ai/api/generate/submit`, then query `GET https://api.vidgo.ai/api/generate/status/{task_id}` with the task ID returned by your submission.

> [accordion] Spring on the Rooftop
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/music",
>   "input": {
>     "composition_plan": {
>       "positive_global_styles": [
>         "indie folk with subtle organic electronica",
>         "96 BPM, 4/4",
>         "warm clear female lead vocal",
>         "acoustic guitar, rounded bass, delicate electronic percussion",
>         "hopeful intimate song about planting a rooftop garden",
>         "clean spacious studio mix, clear Mandarin and English diction"
>       ],
>       "negative_global_styles": [
>         "harsh distortion",
>         "crowded arrangement",
>         "spoken narration",
>         "abrupt cut off"
>       ],
>       "sections": [
>         {
>           "section_name": "Acoustic introduction",
>           "duration": 8,
>           "positive_local_styles": [
>             "instrumental only",
>             "fingerpicked acoustic guitar introduces a memorable gentle motif",
>             "soft airy pad, no drums"
>           ],
>           "negative_local_styles": [],
>           "lines": []
>         },
>         {
>           "section_name": "Mandarin verse",
>           "duration": 16,
>           "positive_local_styles": [
>             "sing the supplied Mandarin lyrics clearly",
>             "intimate gentle lead vocal",
>             "fingerpicked guitar and soft bass, restrained percussion"
>           ],
>           "negative_local_styles": [],
>           "lines": [
>             "把春天种在屋顶上",
>             "等风翻过旧砖墙"
>           ]
>         },
>         {
>           "section_name": "English chorus",
>           "duration": 16,
>           "positive_local_styles": [
>             "sing the supplied English lyrics clearly",
>             "open uplifting melody",
>             "full warm drums and bass enter",
>             "layered backing vocal harmonies, retain the acoustic motif"
>           ],
>           "negative_local_styles": [],
>           "lines": [
>             "Little seeds above the street",
>             "Find the sunlight, find the beat"
>           ]
>         },
>         {
>           "section_name": "Instrumental resolution",
>           "duration": 8,
>           "positive_local_styles": [
>             "instrumental only, no vocals",
>             "return to fingerpicked guitar motif",
>             "drums fall away, warm sustained final chord with natural decay"
>           ],
>           "negative_local_styles": [],
>           "lines": []
>         }
>       ]
>     },
>     "respect_sections_durations": true,
>     "output_format": "mp3_44100_128"
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "W6KDNIQQOUAE1K8V",
>     "status": "finished",
>     "created_time": "2026-09-27T10:48:12",
>     "progress": 100,
>     "error_message": null,
>     "files": [
>       {
>         "file_type": "audio",
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/music/text-to-audio/v1/01/output.mp3"
>       }
>     ]
>   }
> }
> ```
>
> <audio aria-label="Spring on the Rooftop" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/music/text-to-audio/v1/01/output.mp3)

> [accordion] The Clockmaker's Clues
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/music",
>   "input": {
>     "text": "Create a 30-second instrumental jazz cue for a puzzle scene in a clockmaker's workshop. 104 BPM, lightly swinging 4/4, playful curiosity with a hint of mystery. Use plucked upright bass, brushed drums, muted trumpet and vibraphone, each clearly separated in a warm intimate acoustic mix. Begin with a distinctive four-note vibraphone motif over bass; let muted trumpet answer and develop that motif in the middle, while brush accents subtly increase the momentum. Resolve with a short, clean ensemble cadence and allow the last note to decay naturally. No vocals, choir, spoken words, ticking sound effects or abrupt cutoff.",
>     "duration": 30,
>     "is_instrumental": true,
>     "output_format": "mp3_44100_128"
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "EBXJLNFGE8WD7114",
>     "status": "finished",
>     "created_time": "2026-09-27T10:48:48",
>     "progress": 100,
>     "error_message": null,
>     "files": [
>       {
>         "file_type": "audio",
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/music/text-to-audio/v1/02/output.mp3"
>       }
>     ]
>   }
> }
> ```
>
> <audio aria-label="The Clockmaker's Clues" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/music/text-to-audio/v1/02/output.mp3)

> [accordion] Light Through the Kelp Forest
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/music",
>   "input": {
>     "text": "Compose a 40-second instrumental score for a nature documentary gliding through a sunlit underwater kelp forest. 72 BPM, gently flowing 4/4. Felt piano, delicate harp harmonics, soft sustained strings and a subtle low ambient texture; luminous, patient, full of quiet wonder. Start sparsely with isolated felt-piano notes and wide space. Gradually introduce harp and strings into an expansive rising harmony through the middle third, suggesting shafts of light moving through long fronds. Gently thin the arrangement and finish on a peaceful resolved chord with natural reverb decay. Detailed soft dynamics, wide stereo depth, no heavy percussion, no vocals or choir, no narration, no water sound effects, no abrupt cutoff.",
>     "duration": 40,
>     "is_instrumental": true,
>     "output_format": "mp3_44100_128"
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "VAK9IGFH57XPEUGP",
>     "status": "finished",
>     "created_time": "2026-09-27T10:49:30",
>     "progress": 100,
>     "error_message": null,
>     "files": [
>       {
>         "file_type": "audio",
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/music/text-to-audio/v1/03/output.mp3"
>       }
>     ]
>   }
> }
> ```
>
> <audio aria-label="Light Through the Kelp Forest" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/music/text-to-audio/v1/03/output.mp3)


# ElevenLabs V3 TTS
Source: https://docs.vidgo.ai/api-manual/music-series/elevenlabs-tts-eleven-v3

api-manual/music-series/elevenlabs-tts-eleven-v3.json POST /api/generate/submit
Turn text into spoken audio with ElevenLabs V3, with selectable voices, language control, and optional timestamps for narration and dialogue.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn text into spoken audio with ElevenLabs V3, with selectable voices, language control, and optional timestamps for narration and dialogue.

## Models

* **elevenlabs/tts/eleven-v3** - Convert text into speech.

## Features

* Convert text into speech.
* Optionally request timestamps alongside the generated audio.
* Select a voice with `voice` and specify the spoken language with `language_code`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Text

* `text`: Required. Supply the text to speak. Length: **1-5,000 characters**.
* Leading and trailing whitespace is trimmed before the character limit is checked.

### Voice

* `voice`: Optional. Select the voice. String. Default: `Rachel`.
* Available values: `Aria`, `Roger`, `Sarah`, `Laura`, `Charlie`, `George`, `Callum`, `River`, `Liam`, `Charlotte`, `Alice`, `Matilda`, `Will`, `Jessica`, `Eric`, `Chris`, `Brian`, `Daniel`, `Lily`, `Bill`, `Rachel`.

### Stability

* `stability`: Optional. Control voice stability. Number from `0` to `1`. Default: `0.5`.

### Timestamps

* `timestamps`: Optional. Request timestamp information with the audio. Use `true` or `false`. Default: `false`.

### Language Code

* `language_code`: Optional. ISO 639-1 language code, such as `en` for English or `zh` for Chinese.

### Apply Text Normalization

* `apply_text_normalization`: Optional. Select text normalization behavior. Accepted values: `auto`, `on`, `off`. Default: `auto`.

## Output Files

* Read every item in the returned `files` array.
* Generated audio uses `file_type: "audio"`.
* Read returned `timestamps.json` as a separate item with `file_type: "other"`. Preserve optional file metadata when present.

## Examples

Each example pairs a request with its completed task response and generated audio. Submit the request to `POST https://api.vidgo.ai/api/generate/submit`, then query `GET https://api.vidgo.ai/api/generate/status/{task_id}` with the task ID returned by your submission.

> [accordion] Stars in the attic
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/tts/eleven-v3",
>   "input": {
>     "text": "[whispers] Wait... that ticking is coming from inside the wall. [curious] This little brass key fits the crack. One turn, and... [excited] the whole ceiling is full of stars! [laughs] Grandpa, you built a planetarium in the attic. [exhales] All these years, it was waiting for us.",
>     "voice": "Charlotte",
>     "language_code": "en",
>     "stability": 0.5,
>     "timestamps": false,
>     "apply_text_normalization": "auto"
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "LPV36KJQ0Y2SYT21",
>     "status": "finished",
>     "files": [
>       {
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/tts/eleven-v3/v1/01/speech.mp3",
>         "file_type": "audio",
>         "file_name": "speech.mp3"
>       }
>     ],
>     "created_time": "2026-09-27T10:45:57",
>     "error_message": null,
>     "progress": 100
>   }
> }
> ```
>
> <audio aria-label="Stars in the attic" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/tts/eleven-v3/v1/01/speech.mp3)

> [accordion] My first dumpling
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/tts/eleven-v3",
>   "input": {
>     "text": "[sighs] 我练了一个下午，面团还是歪歪扭扭的。奶奶看了一眼，说：别急，手放轻一点。[curious] 这样吗？先对折，再慢慢捏紧……[laughs] 真的站住了！虽然像只胖企鹅，但这是我第一次包好饺子。原来，有些本事不是听懂的，是一起做着学会的。",
>     "voice": "Rachel",
>     "language_code": "zh",
>     "stability": 0.5,
>     "timestamps": false,
>     "apply_text_normalization": "auto"
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "VLXQ5DI7PNWFRJQ4",
>     "status": "finished",
>     "files": [
>       {
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/tts/eleven-v3/v1/02/speech.mp3",
>         "file_type": "audio",
>         "file_name": "speech.mp3"
>       }
>     ],
>     "created_time": "2026-09-27T10:46:38",
>     "error_message": null,
>     "progress": 100
>   }
> }
> ```
>
> <audio aria-label="My first dumpling" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/tts/eleven-v3/v1/02/speech.mp3)

> [accordion] The character of sound
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/tts/eleven-v3",
>   "input": {
>     "text": "[curious] Why does a violin sound different from a flute when both play the same note? Their fundamental frequency can match, but the overtones do not. [excited] Those extra vibrations give each instrument its own sound. At 440 Hz, the pitch is A; the pattern above it is what makes the voice of the instrument recognizable.",
>     "voice": "George",
>     "language_code": "en",
>     "stability": 0.5,
>     "timestamps": true,
>     "apply_text_normalization": "on"
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "0E09515629QS87GR",
>     "status": "finished",
>     "files": [
>       {
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/tts/eleven-v3/v1/03/speech.mp3",
>         "file_type": "audio",
>         "file_name": "speech.mp3"
>       },
>       {
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/tts/eleven-v3/v1/03/timestamps.json",
>         "file_type": "other",
>         "file_name": "timestamps.json"
>       }
>     ],
>     "created_time": "2026-09-27T10:47:13",
>     "error_message": null,
>     "progress": 100
>   }
> }
> ```
>
> <audio aria-label="The character of sound" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/tts/eleven-v3/v1/03/speech.mp3)
>
> [Download timestamps.json](https://cdn.vidgo.ai/apis/models/elevenlabs/tts/eleven-v3/v1/03/timestamps.json)


# ElevenLabs V4 TTS
Source: https://docs.vidgo.ai/api-manual/music-series/elevenlabs-tts-eleven-v4

api-manual/music-series/elevenlabs-tts-eleven-v4.json POST /api/generate/submit
The ElevenLabs V4 TTS API converts up to 5,000 characters into expressive speech with audio tags, voice and stability controls, 19 formats and timestamps.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

ElevenLabs V4 TTS converts text into expressive speech with Eleven v4, the most emotive ElevenLabs model. Inline audio tags such as `[whispers]` or `[laughs]` direct the delivery, and voice, stability, language and output-format controls cover audiobooks, video voice-overs, podcasts, game characters and e-learning narration in 90+ languages.

Eleven v4 and Eleven v4 Turbo are ElevenLabs text-to-speech models released together in September 2026. They take the same request fields on Vidgo: [ElevenLabs V4 TTS](/api-manual/music-series/elevenlabs-tts-eleven-v4) is the most expressive option, and [ElevenLabs V4 Turbo TTS](/api-manual/music-series/elevenlabs-tts-eleven-v4-turbo) is the faster, lower-cost variant.

## Models

* **elevenlabs/tts/eleven-v4** — Eleven v4 text-to-speech, the most expressive ElevenLabs voice model.

## Features

* Text-to-speech for up to 5,000 characters per request.
* Inline audio tags in `text` for emotion, pauses and non-verbal sounds.
* Any ElevenLabs voice by name or voice ID, with `stability` and `similarity_boost` tuning.
* 90+ languages, with an optional `language_code` hint.
* 19 output formats across MP3, PCM, μ-law, A-law and Opus.
* Optional timestamps for subtitles and lip-sync.

## Required parameters

* `text` (string): The text to speak, 1–5,000 characters, not blank. Audio tags in square brackets, such as `[whispers]`, are allowed.

## Optional parameters

* `voice` (string): ElevenLabs voice name or voice ID. Default `Rachel`. See [Voice Presets](#voice-presets).
* `stability` (number): `0`–`1`. Default `0.5`. Lower values are more expressive, higher values more consistent.
* `similarity_boost` (number): `0`–`1`. Default `0.75`. How closely the output follows the selected voice.
* `output_format` (string): One of 19 MP3, PCM, μ-law, A-law and Opus formats. Default `mp3_44100_128`. See [Output Formats](#output-formats).
* `apply_text_normalization` (string): `auto`, `on` or `off`. Default `auto`. Controls whether numbers, dates and abbreviations are spelled out.
* `timestamps` (boolean): Also return timing data for the speech. Default `false`.
* `language_code` (string): ISO 639-1 language hint, such as `en`, `zh` or `ja`. Omit it to let the model detect the language.

## Output

The finished task returns the speech as an audio file in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload. With `timestamps: true`, the timing data is returned as an additional item in the same array.

## Output Formats

| Family | Values |
| - | - |
| MP3 | `mp3_22050_32`, `mp3_44100_32`, `mp3_44100_64`, `mp3_44100_96`, `mp3_44100_128` (default), `mp3_44100_192` |
| PCM | `pcm_8000`, `pcm_16000`, `pcm_22050`, `pcm_24000`, `pcm_44100`, `pcm_48000` |
| Telephony | `ulaw_8000`, `alaw_8000` |
| Opus | `opus_48000_32`, `opus_48000_64`, `opus_48000_96`, `opus_48000_128`, `opus_48000_192` |

The value reads as codec, sample rate in Hz and, for compressed formats, bitrate in kbps. `pcm_*` suits further audio processing; `ulaw_8000` and `alaw_8000` suit phone systems.

## Voice Presets

`voice` accepts any ElevenLabs voice name or voice ID available to the model. Commonly used names include `Aria`, `Roger`, `Sarah`, `Laura`, `Charlie`, `George`, `Callum`, `River`, `Liam`, `Charlotte`, `Alice`, `Matilda`, `Will`, `Jessica`, `Eric`, `Chris`, `Brian`, `Daniel`, `Lily`, `Bill`, `Rachel`. The default is `Rachel`.

## Examples

* **Story narration with audio tags**: `[whispers]`, `[curious]`, `[excited]` and `[laughs]` inside `text`, with `voice: Charlotte` and a lower `stability` for more expressive delivery.
* **Podcast intro as WAV-ready PCM with timestamps**: `output_format: pcm_44100`, `timestamps: true` for subtitle alignment, and a `callback_url`.

## Notes

* Credits are reserved at submission and depend on the number of characters in `text` as sent, including spaces, line breaks and audio tags. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* `text` is limited to 5,000 characters per request. Split longer scripts into several tasks.
* This endpoint produces a single voice per request. Multi-speaker dialogue, voice cloning, streaming, `seed` and `speed` are not available through it; unsupported fields return 400.
* For lower cost and faster turnaround with the same fields, use [ElevenLabs V4 Turbo TTS](/api-manual/music-series/elevenlabs-tts-eleven-v4-turbo).


# ElevenLabs V4 Turbo TTS
Source: https://docs.vidgo.ai/api-manual/music-series/elevenlabs-tts-eleven-v4-turbo

api-manual/music-series/elevenlabs-tts-eleven-v4-turbo.json POST /api/generate/submit
The ElevenLabs V4 Turbo TTS API turns up to 5,000 characters into speech with audio tags, 90+ languages, MP3/PCM/Opus output and optional timestamps.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

ElevenLabs V4 Turbo TTS is the faster, lower-cost Eleven v4 text-to-speech model. It reads up to 5,000 characters per request in 90+ languages with the same voice, stability, audio-tag and output-format controls as Eleven v4, so it fits high-volume narration, app voice prompts, notifications and drafts you may later re-render with Eleven v4.

Eleven v4 and Eleven v4 Turbo are ElevenLabs text-to-speech models released together in September 2026. They take the same request fields on Vidgo: [ElevenLabs V4 TTS](/api-manual/music-series/elevenlabs-tts-eleven-v4) is the most expressive option, and [ElevenLabs V4 Turbo TTS](/api-manual/music-series/elevenlabs-tts-eleven-v4-turbo) is the faster, lower-cost variant.

## Models

* **elevenlabs/tts/eleven-v4-turbo** — Eleven v4 Turbo text-to-speech, the faster and lower-cost variant.

## Features

* Text-to-speech for up to 5,000 characters per request.
* Inline audio tags in `text` for emotion, pauses and non-verbal sounds.
* Any ElevenLabs voice by name or voice ID, with `stability` and `similarity_boost` tuning.
* 90+ languages, with an optional `language_code` hint.
* 19 output formats across MP3, PCM, μ-law, A-law and Opus.
* Optional timestamps for subtitles and lip-sync.

## Required parameters

* `text` (string): The text to speak, 1–5,000 characters, not blank. Audio tags in square brackets, such as `[whispers]`, are allowed.

## Optional parameters

* `voice` (string): ElevenLabs voice name or voice ID. Default `Rachel`. See [Voice Presets](#voice-presets).
* `stability` (number): `0`–`1`. Default `0.5`. Lower values are more expressive, higher values more consistent.
* `similarity_boost` (number): `0`–`1`. Default `0.75`. How closely the output follows the selected voice.
* `output_format` (string): One of 19 MP3, PCM, μ-law, A-law and Opus formats. Default `mp3_44100_128`. See [Output Formats](#output-formats).
* `apply_text_normalization` (string): `auto`, `on` or `off`. Default `auto`. Controls whether numbers, dates and abbreviations are spelled out.
* `timestamps` (boolean): Also return timing data for the speech. Default `false`.
* `language_code` (string): ISO 639-1 language hint, such as `en`, `zh` or `ja`. Omit it to let the model detect the language.

## Output

The finished task returns the speech as an audio file in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload. With `timestamps: true`, the timing data is returned as an additional item in the same array.

## Output Formats

| Family | Values |
| - | - |
| MP3 | `mp3_22050_32`, `mp3_44100_32`, `mp3_44100_64`, `mp3_44100_96`, `mp3_44100_128` (default), `mp3_44100_192` |
| PCM | `pcm_8000`, `pcm_16000`, `pcm_22050`, `pcm_24000`, `pcm_44100`, `pcm_48000` |
| Telephony | `ulaw_8000`, `alaw_8000` |
| Opus | `opus_48000_32`, `opus_48000_64`, `opus_48000_96`, `opus_48000_128`, `opus_48000_192` |

The value reads as codec, sample rate in Hz and, for compressed formats, bitrate in kbps. `pcm_*` suits further audio processing; `ulaw_8000` and `alaw_8000` suit phone systems.

## Voice Presets

`voice` accepts any ElevenLabs voice name or voice ID available to the model. Commonly used names include `Aria`, `Roger`, `Sarah`, `Laura`, `Charlie`, `George`, `Callum`, `River`, `Liam`, `Charlotte`, `Alice`, `Matilda`, `Will`, `Jessica`, `Eric`, `Chris`, `Brian`, `Daniel`, `Lily`, `Bill`, `Rachel`. The default is `Rachel`.

## Examples

* **Notification prompt**: a short transactional message with `voice: Brian` and the default MP3 output.
* **French narration as Opus**: `language_code: fr`, `output_format: opus_48000_64` for compact app audio, with a `callback_url`.

## Notes

* Credits are reserved at submission and depend on the number of characters in `text` as sent, including spaces, line breaks and audio tags. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* `text` is limited to 5,000 characters per request. Split longer scripts into several tasks.
* This endpoint produces a single voice per request. Multi-speaker dialogue, voice cloning, streaming, `seed` and `speed` are not available through it; unsupported fields return 400.
* Turbo is billed at a lower rate than [ElevenLabs V4 TTS](/api-manual/music-series/elevenlabs-tts-eleven-v4); the request body is identical, so you can switch by changing `model`.


# ElevenLabs TTS Turbo v2.5
Source: https://docs.vidgo.ai/api-manual/music-series/elevenlabs-tts-turbo-v2-5

api-manual/music-series/elevenlabs-tts-turbo-v2-5.json POST /api/generate/submit
Create spoken narration with ElevenLabs Turbo v2.5, shaping pacing, vocal style, and voice similarity with optional text alignment.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create spoken narration with ElevenLabs Turbo v2.5, shaping pacing, vocal style, and voice similarity with optional text alignment.

## Models

* **elevenlabs/tts/turbo-v2.5** - Convert text into speech.

## Features

* Convert text into speech.
* Optionally request timestamps alongside the generated audio.
* Select a voice with `voice` and specify the spoken language with `language_code`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

The parameters below define the `input` object for this model.

### Voice

* `voice`: Optional. Select a voice name or compatible voice ID. String. Default: `Rachel`.

### Stability

* `stability`: Optional. Control voice stability. Number from `0` to `1`. Default: `0.5`.

### Similarity Boost

* `similarity_boost`: Optional. Control voice similarity. Number from `0` to `1`. Default: `0.75`.

### Style

* `style`: Optional. Set the voice style strength. Number from `0` to `1`. Default: `0`.

### Speed

* `speed`: Optional. Set the speech speed. Number from `0.7` to `1.2`. Default: `1`.

### Timestamps

* `timestamps`: Optional. Request timestamp information with the audio. Use `true` or `false`. Default: `false`.

### Previous Text

* `previous_text`: Optional. Supply preceding text as speech context. String. Also accepts `null`.

### Next Text

* `next_text`: Optional. Supply following text as speech context. String. Also accepts `null`.

### Language Code

* `language_code`: Optional. Set the spoken language with an ISO 639-1 code, such as `en`, `fr`, `de`, `ja`, `vi`, `hu`, or `no`. Also accepts `null`.

### Apply Text Normalization

* `apply_text_normalization`: Optional. Control how numbers, dates, and currencies are prepared for speech: `auto` selects normalization automatically, `on` enables it, and `off` reads the supplied text directly. Default: `auto`.

### Text

* `text`: Required. Supply the text to speak. Length: **1-40,000 Unicode characters** after trimming leading and trailing whitespace. Include at least one visible text character.

## Output Files

* Read every item in the returned `files` array.
* Generated audio uses `file_type: "audio"`.
* Timestamp and alignment files may use `file_type: "other"`. Preserve optional file metadata when present.

## Examples

Each example pairs a request with its completed task response and generated audio. Submit the request to `POST https://api.vidgo.ai/api/generate/submit`, then query `GET https://api.vidgo.ai/api/generate/status/{task_id}` with the task ID returned by your submission.

> [accordion] One fold at a time
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/tts/turbo-v2.5",
>   "input": {
>     "similarity_boost": 0.75,
>     "style": 0,
>     "apply_text_normalization": "auto",
>     "text": "Almost there. Fold the top corner down until it meets the center crease. Now press along the edge with your fingertip. That's it. Turn the paper over, and we'll make the next fold together.",
>     "voice": "Rachel",
>     "language_code": "en",
>     "speed": 1.05,
>     "stability": 0.5,
>     "timestamps": false
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "QKTPQMIP4AZY1LGE",
>     "status": "finished",
>     "files": [
>       {
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/tts/turbo-v2.5/v1/01/output.mp3",
>         "file_type": "audio"
>       }
>     ],
>     "created_time": "2026-09-27T10:51:18",
>     "error_message": null,
>     "progress": 100
>   }
> }
> ```
>
> <audio aria-label="One fold at a time" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/tts/turbo-v2.5/v1/01/output.mp3)

> [accordion] Finding north in the night sky
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/tts/turbo-v2.5",
>   "input": {
>     "similarity_boost": 0.75,
>     "style": 0,
>     "apply_text_normalization": "auto",
>     "text": "今晚，我们先找到北斗七星，再从勺口两颗星的方向寻找北极星。它看起来并不耀眼，却能帮我们辨认北方。在黑暗中安静地等待片刻，你会发现，夜空远比第一眼看到的更丰富。",
>     "voice": "Aria",
>     "language_code": "zh",
>     "speed": 0.95,
>     "stability": 0.6,
>     "timestamps": true
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "2OAYYBTX2HQT9R8P",
>     "status": "finished",
>     "files": [
>       {
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/tts/turbo-v2.5/v1/02/output.mp3",
>         "file_type": "audio"
>       },
>       {
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/tts/turbo-v2.5/v1/02/attachment-1.json",
>         "file_type": "other"
>       }
>     ],
>     "created_time": "2026-09-27T10:54:18",
>     "error_message": null,
>     "progress": 100
>   }
> }
> ```
>
> <audio aria-label="Finding north in the night sky" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/tts/turbo-v2.5/v1/02/output.mp3)
>
> [Download attachment-1.json](https://cdn.vidgo.ai/apis/models/elevenlabs/tts/turbo-v2.5/v1/02/attachment-1.json)

> [accordion] Two cups of tea
>
> ```json Request theme={null}
> {
>   "model": "elevenlabs/tts/turbo-v2.5",
>   "input": {
>     "similarity_boost": 0.75,
>     "style": 0,
>     "apply_text_normalization": "auto",
>     "text": "Hoy volví a la casa de mi abuela. En la cocina seguía el reloj de siempre, cinco minutos adelantado. Preparé dos tazas de té por costumbre y, al darme cuenta, sonreí. Algunos gestos también guardan recuerdos.",
>     "voice": "Sarah",
>     "language_code": "es",
>     "speed": 0.98,
>     "stability": 0.45,
>     "timestamps": false,
>     "previous_text": "Hacía años que no cruzaba aquella puerta.",
>     "next_text": "Después abrí las ventanas para dejar entrar la luz."
>   }
> }
> ```
>
> ```json Completed task response theme={null}
> {
>   "code": 200,
>   "data": {
>     "task_id": "KAQOQIL83QO5TWA6",
>     "status": "finished",
>     "files": [
>       {
>         "file_url": "https://cdn.vidgo.ai/apis/models/elevenlabs/tts/turbo-v2.5/v1/03/output.mp3",
>         "file_type": "audio"
>       }
>     ],
>     "created_time": "2026-09-27T10:52:07",
>     "error_message": null,
>     "progress": 100
>   }
> }
> ```
>
> <audio aria-label="Two cups of tea" />
>
> [Download audio](https://cdn.vidgo.ai/apis/models/elevenlabs/tts/turbo-v2.5/v1/03/output.mp3)


# Extend Music
Source: https://docs.vidgo.ai/api-manual/music-series/extend-music

api-manual/music-series/extend-music.json POST /api/generate/submit
Extend or modify existing music by creating a continuation based on a source audio track

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint extends an existing audio track from a specified point
* You can customize the extension with new style, prompt, and title, or inherit from the original track
* Use `default_param_flag` to control whether to use custom parameters or original track settings

## Parameter Details

* In Custom Mode ( `default_param_flag: true` ):

  * `prompt`, `style`, `title`, and `continue_at` are **required**

  * `continue_at` specifies the timestamp (in seconds) where extension begins

  * Character limits vary by model:
    * **V4**: `prompt` 3000 characters, `style` 200 characters, `title` 80 characters
    * **V4\_5 & V4\_5PLUS**: `prompt` 5000 characters, `style` 1000 characters, `title` 100 characters
    * **V4\_5ALL**: `prompt` 5000 characters, `style` 1000 characters, `title` 80 characters
    * **V5**: `prompt` 5000 characters, `style` 1000 characters, `title` 100 characters

* In Simple Mode ( `default_param_flag: false` ):

  * Only `audio_id` and `mv` are required

## Usage Notes

* Ensure you have a valid `audio_id` from a previously generated track before using this endpoint

* The extension will seamlessly continue from the specified `continue_at` timestamp

## Optional parameters

* `negative_tags` (string): Music styles or characteristics to exclude from the extension.

* `vocal_gender` (string): Vocal gender preference. Use `m` for male, `f` for female. Note: This parameter increases the probability but cannot guarantee adherence.

* `style_weight` (number): Strength of adherence to style. Range 0-1, up to 2 decimals.

* `weirdness_constraint` (number): Controls creative deviation. Range 0-1, up to 2 decimals.

* `audio_weight` (number): Balance weight for audio features. Range 0-1, up to 2 decimals.

* `persona_id` (string): Persona ID to apply to the extended music. How to generate persona\_id, visit [generate-persona](/api-manual/music-series/generate-persona).


# Generate Lyrics
Source: https://docs.vidgo.ai/api-manual/music-series/generate-lyrics

api-manual/music-series/generate-lyrics.json POST /api/generate/submit
AI-powered lyrics generation based on themes and descriptions

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint generates song lyrics based on your theme or description
* Multiple lyric variations may be generated for each request
* Use the generated lyrics with the Generate Music endpoint in Custom Mode

## Parameter Details

* `prompt` (required): Description of the desired lyrics content
  * Be specific about theme, mood, style, or story elements
  * Maximum length: 200 words
  * Include details like genre feel, emotional tone, and narrative elements


# Generate Music
Source: https://docs.vidgo.ai/api-manual/music-series/generate-music

api-manual/music-series/generate-music.json POST /api/generate/submit
AI-powered music generation with customizable styles and vocals

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint creates music based on your text prompt
* Multiple variations will be generated for each request
* You can control detail level with custom mode and instrumental settings

## Parameter Details

* In Custom Mode ( `custom_mode: true` ):

  * If `instrumental: true` : `style` and `title` are required

  * If `instrumental: false` : `style`, `prompt`, and `title` are required

  * Character limits vary by model:

    * **V4**: `prompt` 3000 characters, `style` 200 characters

    * **V4\_5 & V4\_5PLUS**: `prompt` 5000 characters, `style` 1000 characters

    * **V4\_5ALL**: `prompt` 5000 characters, `style` 1000 characters

    * **V5**: `prompt` 5000 characters, `style` 1000 characters

  * `title` length limit: 80 characters (all models)

* In Non-custom Mode ( `custom_mode: false` ):

  * `prompt` length limit: 500 characters

  * Other parameters should be left empty

## Usage Notes

* Recommendation for new users: Start with `custom_mode: false` for simpler usage

## Optional parameters

* `vocal_gender` (string): Vocal gender preference. Use `m` for male, `f` for female. Note: This parameter is only effective when `custom_mode` is `true` . Based on practice, this parameter can only increase the probability but cannot guarantee adherence to male/female voice instructions.

* `style_weight` (number): Strength of adherence to style. Range 0-1, up to 2 decimals. Example: `0.65` .

* `weirdness_constraint` (number): Controls creative deviation. Range 0-1, up to 2 decimals. Example: `0.65` .

* `audio_weight` (number): Balance weight for audio features. Range 0-1, up to 2 decimals. Example: `0.65` .

* `persona_id` (string): Persona ID to apply to the generated music. Use this to apply a specific persona style to your music generation. Only available when Custom Mode is enabled. How to generate persona\_id, visit [generate-persona](/api-manual/music-series/generate-persona).


# Generate Music Cover
Source: https://docs.vidgo.ai/api-manual/music-series/generate-music-cover

api-manual/music-series/generate-music-cover.json POST /api/generate/submit
The Generate Music Cover API creates cover artwork images for a finished Suno music task from its task_id; each task can generate covers only once.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint creates cover artwork images for an existing music track
* Requires a valid task\_id from a previously generated music task
* The generated images are returned in the `generate_cover` field of the task result

## Parameter Details

* `task_id` (required): The task ID from a completed music generation task
  * Must be a valid task\_id returned from the Generate Music or Extend Music endpoints
  * The original task must have completed successfully

## Usage Notes

* Cover images can only be generated once per original task
* Results are delivered via the callback URL when processing is complete


# Generate Persona
Source: https://docs.vidgo.ai/api-manual/music-series/generate-persona

api-manual/music-series/generate-persona.json POST /api/generate/submit
Create reusable musical personas from existing audio tracks

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint creates a reusable musical persona from an existing audio track
* Personas capture the vocal style, genre characteristics, and musical personality
* Once created, personas can be used with the Generate Music endpoint via the `persona_id` parameter

## Parameter Details

* `task_id` (required): Task ID from a completed music generation ([Generate](/api-manual/music-series/generate-music), [Extend](/api-manual/music-series/extend-music), [Upload Cover](/api-manual/music-series/upload-and-cover-audio), or [Upload Extend](/api-manual/music-series/upload-and-extend-audio))
* `audio_id` (required): Specific audio track identifier from the callback data
* `name` (required): A descriptive name that captures the musical style or character
* `description` (required): Detailed description including:
  * Genre and mood
  * Instrumentation preferences
  * Vocal characteristics
  * Unique musical personality traits

## Usage Notes

* Each audio track can only have one persona created from it
* If a persona already exists for the audio, the API returns error code 409
* You can apply the `persona_id` to the following endpoints:
  * [Generate Music](/api-manual/music-series/generate-music)
  * [Extend Music](/api-manual/music-series/extend-music)
  * [Upload And Cover Audio](/api-manual/music-series/upload-and-cover-audio)
  * [Upload And Extend Audio](/api-manual/music-series/upload-and-extend-audio)
* Provide detailed descriptions for best results when using the persona


# Get Timestamped Lyrics
Source: https://docs.vidgo.ai/api-manual/music-series/get-timestamped-lyrics

api-manual/music-series/get-timestamped-lyrics.json POST /api/generate/submit
Retrieve synchronized lyrics with precise timestamps

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint retrieves synchronized lyrics with precise timestamps for generated audio tracks
* Use this to create karaoke-style displays, subtitles, or lyric visualizations
* Requires a completed music generation task with vocals

## Parameter Details

* Required parameters:

  * `task_id`: The unique identifier from a previous music generation task

  * `audio_id`: The specific audio track identifier from the task result

* Both identifiers are obtained from the response of music generation endpoints or their callbacks

## Usage Notes

* This endpoint only works with audio tracks that contain vocals
* The `hoot_cer` (Character Error Rate) value indicates alignment precision - lower values mean better accuracy
* Use `waveform_data` for creating audio visualizations alongside the lyrics

## Response Fields

* `aligned_words` (array): List of lyric words with timing information
  * `word` (string): The lyric text, may include section markers like `[Verse]`, `[Chorus]`
  * `start_s` (number): Word start time in seconds
  * `end_s` (number): Word end time in seconds
  * `success` (boolean): Whether the word was successfully aligned
  * `palign` (number): Alignment confidence score

* `waveform_data` (array): Numerical data for audio waveform visualization

* `hoot_cer` (number): Alignment precision score (Character Error Rate)

* `is_streamed` (boolean): Indicates if the audio is a streamed track


# Gemini 3.1 Flash TTS
Source: https://docs.vidgo.ai/api-manual/music-series/google-gemini-3-1-flash-text-to-speech

api-manual/music-series/google-gemini-3-1-flash-text-to-speech.json POST /api/generate/submit
Generate speech with Gemini 3.1 Flash TTS, using style instructions and selectable voices for narration or dialogue with two speakers.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate speech with Gemini 3.1 Flash TTS, using style instructions and selectable voices for narration or dialogue with two speakers.

## Models

* **google/gemini-3.1-flash/text-to-speech** - Convert text into speech.

## Features

* Convert text into speech.
* Use one voice or two speaker aliases.
* Control delivery with `style_instructions` and `temperature`.
* Select a voice with `voice` and specify the spoken language with `language_code`.

## Output Options

### Output Format

* `output_format`: Optional. Accepted values: `mp3`, `wav`, `ogg_opus`. Default: `mp3`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Use the fields in the parameter reference to configure speech generation.

* When supplying `speakers`, provide exactly two distinct `speaker_id` aliases using only letters, digits, and underscores. Speech uses the voice assigned to each speaker.

### Text

* `text`: Required. Supply the text to speak. Length: **1-50,000 characters**.
* Include the words to speak, with optional audio tags such as `[short pause]`, `[whispering]`, and `[laughing]`. For dialogue, prefix each turn with its matching `speaker_id` followed by a colon.

### Style Instructions

* `style_instructions`: Optional. Describe how the speech should be delivered.

### Voice

* `voice`: Optional. Select the voice. String. Default: `Kore`.

> [accordion] Supported voice values
>
> * `Achernar`
> * `Achird`
> * `Algenib`
> * `Algieba`
> * `Alnilam`
> * `Aoede`
> * `Autonoe`
> * `Callirrhoe`
> * `Charon`
> * `Despina`
> * `Enceladus`
> * `Erinome`
> * `Fenrir`
> * `Gacrux`
> * `Iapetus`
> * `Kore`
> * `Laomedeia`
> * `Leda`
> * `Orus`
> * `Pulcherrima`
> * `Puck`
> * `Rasalgethi`
> * `Sadachbia`
> * `Sadaltager`
> * `Schedar`
> * `Sulafat`
> * `Umbriel`
> * `Vindemiatrix`
> * `Zephyr`
> * `Zubenelgenubi`

### Language Code

* `language_code`: Optional. Select a value below, or omit it to detect the language from `text`.

> [accordion] Supported language code values
>
> * `Arabic (Egypt)`
> * `Bangla (Bangladesh)`
> * `Dutch (Netherlands)`
> * `English (India)`
> * `English (US)`
> * `French (France)`
> * `German (Germany)`
> * `Hindi (India)`
> * `Indonesian (Indonesia)`
> * `Italian (Italy)`
> * `Japanese (Japan)`
> * `Korean (South Korea)`
> * `Marathi (India)`
> * `Polish (Poland)`
> * `Portuguese (Brazil)`
> * `Romanian (Romania)`
> * `Russian (Russia)`
> * `Spanish (Spain)`
> * `Tamil (India)`
> * `Telugu (India)`
> * `Thai (Thailand)`
> * `Turkish (Turkey)`
> * `Ukrainian (Ukraine)`
> * `Vietnamese (Vietnam)`
> * `Afrikaans (South Africa)`
> * `Albanian (Albania)`
> * `Amharic (Ethiopia)`
> * `Arabic (World)`
> * `Armenian (Armenia)`
> * `Azerbaijani (Azerbaijan)`
> * `Basque (Spain)`
> * `Belarusian (Belarus)`
> * `Bulgarian (Bulgaria)`
> * `Burmese (Myanmar)`
> * `Catalan (Spain)`
> * `Cebuano (Philippines)`
> * `Chinese Mandarin (China)`
> * `Chinese Mandarin (Taiwan)`
> * `Croatian (Croatia)`
> * `Czech (Czech Republic)`
> * `Danish (Denmark)`
> * `English (Australia)`
> * `English (UK)`
> * `Estonian (Estonia)`
> * `Filipino (Philippines)`
> * `Finnish (Finland)`
> * `French (Canada)`
> * `Galician (Spain)`
> * `Georgian (Georgia)`
> * `Greek (Greece)`
> * `Gujarati (India)`
> * `Haitian Creole (Haiti)`
> * `Hebrew (Israel)`
> * `Hungarian (Hungary)`
> * `Icelandic (Iceland)`
> * `Javanese (Java)`
> * `Kannada (India)`
> * `Konkani (India)`
> * `Lao (Laos)`
> * `Latin (Vatican City)`
> * `Latvian (Latvia)`
> * `Lithuanian (Lithuania)`
> * `Luxembourgish (Luxembourg)`
> * `Macedonian (North Macedonia)`
> * `Maithili (India)`
> * `Malagasy (Madagascar)`
> * `Malay (Malaysia)`
> * `Malayalam (India)`
> * `Mongolian (Mongolia)`
> * `Nepali (Nepal)`
> * `Norwegian Bokmal (Norway)`
> * `Norwegian Nynorsk (Norway)`
> * `Odia (India)`
> * `Pashto (Afghanistan)`
> * `Persian (Iran)`
> * `Portuguese (Portugal)`
> * `Punjabi (India)`
> * `Serbian (Serbia)`
> * `Sindhi (India)`
> * `Sinhala (Sri Lanka)`
> * `Slovak (Slovakia)`
> * `Slovenian (Slovenia)`
> * `Spanish (Latin America)`
> * `Spanish (Mexico)`
> * `Swahili (Kenya)`
> * `Swedish (Sweden)`
> * `Urdu (Pakistan)`

### Speakers

* `speakers`: Optional. Supply exactly 2 items. Each item is an object.

* `speakers[].voice`: Required. Select a voice for this speaker.

> [accordion] Supported voice values
>
> * `Achernar`
> * `Achird`
> * `Algenib`
> * `Algieba`
> * `Alnilam`
> * `Aoede`
> * `Autonoe`
> * `Callirrhoe`
> * `Charon`
> * `Despina`
> * `Enceladus`
> * `Erinome`
> * `Fenrir`
> * `Gacrux`
> * `Iapetus`
> * `Kore`
> * `Laomedeia`
> * `Leda`
> * `Orus`
> * `Pulcherrima`
> * `Puck`
> * `Rasalgethi`
> * `Sadachbia`
> * `Sadaltager`
> * `Schedar`
> * `Sulafat`
> * `Umbriel`
> * `Vindemiatrix`
> * `Zephyr`
> * `Zubenelgenubi`

* `speakers[].speaker_id`: Required. String. Use only letters, digits, and underscores.

### Temperature

* `temperature`: Optional. Control speech-generation variability. Number from `0` to `2`. Default: `1`.

## Output Files

* Read every item in the returned `files` array.
* Generated audio uses `file_type: "audio"`.
* Read the audio URL from `files[].file_url`.


# Gemini 3.8 Flash Lite TTS
Source: https://docs.vidgo.ai/api-manual/music-series/google-gemini-3-8-flash-lite-text-to-speech

api-manual/music-series/google-gemini-3-8-flash-lite-text-to-speech.json POST /api/generate/submit
The Gemini 3.8 Flash-Lite TTS API generates WAV speech at scale from a single voice or two-speaker dialogue, with 30 preset voices and style instructions.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Gemini 3.8 Flash-Lite TTS is Google's high-throughput text-to-speech model. It turns text, or a scripted conversation between two speakers, into WAV audio with 30 preset voices, auto-detects the language (Google lists 101) and takes plain-language style instructions. Use it for read-aloud features, notifications, voice-agent prompts and other high-volume, everyday speech.

Gemini 3.8 Flash-Lite TTS takes the same request fields as [Gemini 3.8 Flash TTS](/api-manual/music-series/google-gemini-3-8-flash-text-to-speech), which Google positions for maximum voice fidelity and complex dialogue. Switch between them by changing `model`.

## Models

* **google/gemini-3.8-flash-lite/text-to-speech** — Gemini 3.8 Flash-Lite text-to-speech for high-volume single-voice or two-speaker audio.

## Features

* Text-to-speech with one voice, or a scripted dialogue between two speakers.
* 30 preset voices, set per request or per speaker.
* Plain-language `style_instructions` for tone, pace, accent and emotion, globally or per dialogue turn.
* Automatic language detection (Google lists 101 languages).
* WAV audio output.

## Input modes

* **Single speaker**: send `text` and, optionally, `voice` (default `Kore`).
* **Two-speaker dialogue**: send exactly two `speakers` and a list of `turns`. Do not send top-level `text` or `voice` in this mode.

## Required parameters

Single speaker:

* `text` (string): The text to speak. Not blank.

Two-speaker dialogue:

* `speakers` (object\[]): Exactly two objects, each with only:
  * `speaker_id` (string): A non-empty alias, unique within the request, such as `host`.
  * `voice` (string): One of the [Voice Presets](#voice-presets).
* `turns` (object\[]): At least one line, in speaking order:
  * `speaker_id` (string): Must match one of the two `speakers`.
  * `text` (string): What that speaker says. Not blank.
  * `style_instructions` (string, optional): Direction for this line only.

## Optional parameters

* `voice` (string): Single-speaker mode only. One of the [Voice Presets](#voice-presets). Default `Kore`.
* `style_instructions` (string): Global direction for the whole request, such as `Calm, warm documentary narration`. Works in both modes.
* `enable_safety_checker` (boolean): Run the upstream safety checker. Default `true`.

## Output

The finished task returns one WAV audio file in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload. The output format is fixed; there is no `output_format` field.

## Voice Presets

`voice` and `speakers[].voice` accept these 30 values (case-sensitive):

> [accordion] Supported voice values
>
> * `Achernar`
> * `Achird`
> * `Algenib`
> * `Algieba`
> * `Alnilam`
> * `Aoede`
> * `Autonoe`
> * `Callirrhoe`
> * `Charon`
> * `Despina`
> * `Enceladus`
> * `Erinome`
> * `Fenrir`
> * `Gacrux`
> * `Iapetus`
> * `Kore`
> * `Laomedeia`
> * `Leda`
> * `Orus`
> * `Pulcherrima`
> * `Puck`
> * `Rasalgethi`
> * `Sadachbia`
> * `Sadaltager`
> * `Schedar`
> * `Sulafat`
> * `Umbriel`
> * `Vindemiatrix`
> * `Zephyr`
> * `Zubenelgenubi`

## Examples

* **Single-speaker read-aloud**: a message summary read with `voice: Puck`.
* **Two-speaker support dialogue**: `agent` (`Kore`) and `caller` (`Charon`) with a global `style_instructions`, plus a `callback_url`.

## Notes

* Credits are reserved at submission and depend on the billable character count: `text`, the global `style_instructions`, and every turn's `text` and `style_instructions`. Speaker IDs are not counted. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* Flash-Lite is billed at a lower rate than [Gemini 3.8 Flash TTS](/api-manual/music-series/google-gemini-3-8-flash-text-to-speech).
* Dialogue supports exactly two speakers. A third speaker, a `speaker_id` in `turns` that is not in `speakers`, or top-level `text`/`voice` in a dialogue request returns 400.
* Put delivery directions in `style_instructions`, not in `text`; the model may read directions inside `text` aloud.
* `output_format`, `language_code`, `temperature`, `seed` and timestamps are not supported on this endpoint, and neither are voice cloning or custom voices.


# Gemini 3.8 Flash TTS
Source: https://docs.vidgo.ai/api-manual/music-series/google-gemini-3-8-flash-text-to-speech

api-manual/music-series/google-gemini-3-8-flash-text-to-speech.json POST /api/generate/submit
The Gemini 3.8 Flash TTS API turns a script or a two-speaker dialogue into WAV audio, with 30 preset voices and style instructions to steer delivery.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Gemini 3.8 Flash TTS is Google's highest-fidelity Gemini 3.8 text-to-speech model. It reads a script with one voice, or performs a two-speaker dialogue, and returns WAV audio. Choose from 30 preset voices, steer tone, pace and emotion with style instructions, and let the model detect the language (Google lists 130). It suits audiobooks, podcast-style conversations, studio narration and character dialogue.

For high-volume, everyday speech with the same request fields, use [Gemini 3.8 Flash-Lite TTS](/api-manual/music-series/google-gemini-3-8-flash-lite-text-to-speech). Switch between them by changing `model`.

## Models

* **google/gemini-3.8-flash/text-to-speech** — Gemini 3.8 Flash text-to-speech for single-voice narration and two-speaker dialogue.

## Features

* Text-to-speech with one voice, or a scripted dialogue between two speakers.
* 30 preset voices, set per request or per speaker.
* Plain-language `style_instructions` for tone, pace, accent and emotion, globally or per dialogue turn.
* Automatic language detection (Google lists 130 languages).
* WAV audio output.

## Input modes

* **Single speaker**: send `text` and, optionally, `voice` (default `Kore`).
* **Two-speaker dialogue**: send exactly two `speakers` and a list of `turns`. Do not send top-level `text` or `voice` in this mode.

## Required parameters

Single speaker:

* `text` (string): The text to speak. Not blank.

Two-speaker dialogue:

* `speakers` (object\[]): Exactly two objects, each with only:
  * `speaker_id` (string): A non-empty alias, unique within the request, such as `host`.
  * `voice` (string): One of the [Voice Presets](#voice-presets).
* `turns` (object\[]): At least one line, in speaking order:
  * `speaker_id` (string): Must match one of the two `speakers`.
  * `text` (string): What that speaker says. Not blank.
  * `style_instructions` (string, optional): Direction for this line only.

## Optional parameters

* `voice` (string): Single-speaker mode only. One of the [Voice Presets](#voice-presets). Default `Kore`.
* `style_instructions` (string): Global direction for the whole request, such as `Calm, warm documentary narration`. Works in both modes.
* `enable_safety_checker` (boolean): Run the upstream safety checker. Default `true`.

## Output

The finished task returns one WAV audio file in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload. The output format is fixed; there is no `output_format` field.

## Voice Presets

`voice` and `speakers[].voice` accept these 30 values (case-sensitive):

> [accordion] Supported voice values
>
> * `Achernar`
> * `Achird`
> * `Algenib`
> * `Algieba`
> * `Alnilam`
> * `Aoede`
> * `Autonoe`
> * `Callirrhoe`
> * `Charon`
> * `Despina`
> * `Enceladus`
> * `Erinome`
> * `Fenrir`
> * `Gacrux`
> * `Iapetus`
> * `Kore`
> * `Laomedeia`
> * `Leda`
> * `Orus`
> * `Pulcherrima`
> * `Puck`
> * `Rasalgethi`
> * `Sadachbia`
> * `Sadaltager`
> * `Schedar`
> * `Sulafat`
> * `Umbriel`
> * `Vindemiatrix`
> * `Zephyr`
> * `Zubenelgenubi`

## Examples

* **Single-speaker narration with style**: `voice: Charon` with a global `style_instructions` for slow documentary delivery.
* **Two-speaker podcast dialogue**: `host` (`Kore`) and `guest` (`Puck`), with per-turn `style_instructions` on two lines and a `callback_url`.

## Notes

* Credits are reserved at submission and depend on the billable character count: `text`, the global `style_instructions`, and every turn's `text` and `style_instructions`. Speaker IDs are not counted. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.
* [Gemini 3.8 Flash-Lite TTS](/api-manual/music-series/google-gemini-3-8-flash-lite-text-to-speech) is billed at a lower rate with the same fields.
* Dialogue supports exactly two speakers. A third speaker, a `speaker_id` in `turns` that is not in `speakers`, or top-level `text`/`voice` in a dialogue request returns 400.
* Put delivery directions in `style_instructions`, not in `text`; the model may read directions inside `text` aloud.
* `output_format`, `language_code`, `temperature`, `seed` and timestamps are not supported on this endpoint, and neither are voice cloning or custom voices.


# MiniMax Music 2.6
Source: https://docs.vidgo.ai/api-manual/music-series/minimax-music-2-6

api-manual/music-series/minimax-music-2-6.json POST /api/generate/submit
Generate songs and instrumental tracks from style prompts, with custom lyrics, automatic lyric generation and configurable audio output.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create complete songs and instrumental tracks from a style prompt. Describe genre, mood, instrumentation, tempo and vocal direction to guide the music.

## Models

Set `model` to `minimax/music-2.6`. Place music parameters inside `input` and the optional `callback_url` at the request root.

## Music Parameters

* `prompt`: Required string of **10-2000 characters** describing the musical direction.
* `lyrics`: Lyrics string of up to **3500 characters**. Supply lyrics when `is_instrumental` and `lyrics_optimizer` are both `false`. You can organize the words with section tags such as `[Verse]` and `[Chorus]`.
* `lyrics_optimizer`: Boolean, default `false`. When `true` and `lyrics` is empty, generates lyrics from `prompt`.
* `is_instrumental`: Boolean, default `false`. Set `true` for instrumental music, keep `lyrics` empty and set `lyrics_optimizer` to `false`.

## Creation Modes

### Custom Lyrics

Write the musical direction in `prompt` and the words to sing in `lyrics`. Keep `is_instrumental` set to `false`.

### Automatic Lyrics

Describe the musical style, song topic and vocal character in `prompt`. Set `lyrics_optimizer` to `true`, keep `lyrics` empty and set `is_instrumental` to `false`.

### Instrumental Music

Describe instruments, rhythm and atmosphere in `prompt`. Set `is_instrumental` to `true`, keep `lyrics` empty and set `lyrics_optimizer` to `false`.

## Audio Settings

`audio_setting` is an optional object with these fields:

* `sample_rate`: Sample rate in Hz. Values: `16000`, `24000`, `32000`, `44100`. Default: `44100`.
* `bitrate`: Bitrate in bits per second. Values: `32000`, `64000`, `128000`, `256000`. Default: `256000`.
* `format`: Audio format. Values: `mp3`, `wav`, `pcm`. Default: `mp3`.

## Retrieve the Audio

Query `GET /api/generate/status/{task_id}` using the identifier from the submission response. Continue checking while the status is `not_started` or `running`.

When the status is `finished`, read the `files` array and use `file_url` from entries with `file_type: "audio"` to retrieve the music. For a `failed` task, read `error_message` and review the request before trying again.


# Music Models
Source: https://docs.vidgo.ai/api-manual/music-series/overview

All music generation models and Suno music actions available on Vidgo API, with their model IDs.

Music models are called through `POST /api/generate/submit`. Send a model ID in `model`, the parameters in `input`, and an optional `callback_url`. The request returns a `task_id` right away and the task runs asynchronously.

## Workflow

Suno and the other providers use different result endpoints:

* **Suno**: poll [Query Music Detail](/api-manual/music-series/query-music-detail), or receive the result at `callback_url` (see [Music Webhook](/api-manual/music-series/music-webhook)).
* **MiniMax, ElevenLabs**: poll the unified [Get Task Status](/api-manual/task-management/status), or receive the result at `callback_url` (see [Webhook Callbacks](/api-manual/task-management/webhooks)).

## Suno

Suno is organized by action. Each action has its own model ID. Start with [Generate Music](/api-manual/music-series/generate-music); several actions extend, cover, or edit a song from an earlier task or an uploaded audio file.

| Action | Model ID |
| - | - |
| [Generate Music](/api-manual/music-series/generate-music) | `generate-music` |
| [Extend Music](/api-manual/music-series/extend-music) | `extend-music` |
| [Upload and Cover Audio](/api-manual/music-series/upload-and-cover-audio) | `upload-and-cover-audio` |
| [Upload and Extend Audio](/api-manual/music-series/upload-and-extend-audio) | `upload-and-extend-audio` |
| [Add Instrumental](/api-manual/music-series/add-instrumental) | `add-instrumental` |
| [Add Vocals](/api-manual/music-series/add-vocals) | `add-vocals` |
| [Get Timestamped Lyrics](/api-manual/music-series/get-timestamped-lyrics) | `get-timestamped-lyrics` |
| [Boost Music Style](/api-manual/music-series/boost-music-style) | `boost-music-style` |
| [Generate Music Cover](/api-manual/music-series/generate-music-cover) | `generate-music-cover` |
| [Replace Section](/api-manual/music-series/replace-section) | `replace-section` |
| [Generate Persona](/api-manual/music-series/generate-persona) | `generate-persona` |
| [Generate Lyrics](/api-manual/music-series/generate-lyrics) | `generate-lyrics` |
| [Convert to WAV](/api-manual/music-series/convert-to-wav) | `convert-to-wav` |
| [Separate Vocals](/api-manual/music-series/vocal-remover/separate-vocals) | `separate-vocals` |
| [Stem Split](/api-manual/music-series/vocal-remover/stem-split) | `stem-split` |
| [Upload and Separate Vocals](/api-manual/music-series/vocal-remover/upload-and-separate-vocals) | `upload-and-separate-vocals` |
| [Create Music Video](/api-manual/music-series/create-music-video) | `create-music-video` |

## Other models

| Model | Provider | Model IDs |
| - | - | - |
| [MiniMax Music 2.6](/api-manual/music-series/minimax-music-2-6) | MiniMax | `minimax/music-2.6` |
| [ElevenLabs Music](/api-manual/music-series/elevenlabs-music) | ElevenLabs | `elevenlabs/music` |

## Notes

* Each model page lists the parameters, limits, and modes for its model IDs. Parameters are not shared across models.
* Generated files are valid for 24 hours after creation. Download and store them promptly.
* URL inputs must be publicly accessible HTTP(S) URLs. To use local files, upload them first with [File Management](/api-manual/file-series/upload-url).


# Replace Section
Source: https://docs.vidgo.ai/api-manual/music-series/replace-section

api-manual/music-series/replace-section.json POST /api/generate/submit
Replace specific sections of generated music tracks

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint replaces a specific time range of an existing music track with new content
* Useful for editing verses, choruses, or any specific part of a generated track
* The replacement content is generated based on your prompt and style tags

## Parameter Details

* `task_id` (required): The original task ID from a completed music generation
* `audio_id` (required): The specific audio track identifier from the callback data
* `prompt` (required): Description of the replacement content (lyrics or musical description)
* `tags` (required): Style tags for the replacement section
* `title` (required): Title for the resulting track
* `infill_start_s` (required): Start time in seconds (minimum 0, up to 2 decimals)
* `infill_end_s` (required): End time in seconds (must be greater than start time)

## Usage Notes

* Ensure `infill_end_s` is always greater than `infill_start_s`
* The time range should be within the duration of the original track
* Use `full_lyrics` when you want to maintain lyrical consistency across the entire track

## Optional parameters

* `negative_tags` (string): Styles to exclude from the replacement section
* `full_lyrics` (string): Complete lyrics for the entire track after modification
* `callback_url` (string): Webhook URL for receiving completion notifications


# Upload and Cover Audio
Source: https://docs.vidgo.ai/api-manual/music-series/upload-and-cover-audio

api-manual/music-series/upload-and-cover-audio.json POST /api/generate/submit
Transform audio tracks into new styles while preserving the original melody

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint transforms an uploaded audio track into a new style while preserving the original melody
* Upload your audio file and specify the desired transformation style
* The original melody and structure are maintained while applying new musical characteristics

## Parameter Details

* **Audio Upload Requirement**:

  * The uploaded audio must not exceed **8 minutes** in length

  * **Note**: For the V4\_5ALL model, the uploaded audio must not exceed **1 minute** in length

  * Provide a valid URL pointing to your audio file via `upload_url`

* In Custom Mode ( `custom_mode: true` ):

  * `style` and `title` are **required**

  * `prompt` is **required** if `instrumental` is `false`

  * Character limits vary by model:
    * **V4**: `prompt` 3000 characters, `style` 200 characters, `title` 80 characters
    * **V4\_5 & V4\_5PLUS**: `prompt` 5000 characters, `style` 1000 characters, `title` 100 characters
    * **V4\_5ALL**: `prompt` 5000 characters, `style` 1000 characters, `title` 80 characters
    * **V5**: `prompt` 5000 characters, `style` 1000 characters, `title` 100 characters

* In Non-custom Mode ( `custom_mode: false` ):

  * `prompt` is **required** to describe the desired transformation (max 500 characters)

  * `style` and `title` should be left empty

## Usage Notes

* Ensure your audio URL is publicly accessible and the file is under 8 minutes (1 minute for V4\_5ALL)

* For best results, use high-quality source audio with clear melody lines

* The `audio_weight` parameter can help balance between preserving the original and applying the new style

## Optional parameters

* `negative_tags` (string): Music styles or characteristics to exclude from the cover.

* `vocal_gender` (string): Vocal gender preference. Use `m` for male, `f` for female. Note: This parameter increases the probability but cannot guarantee adherence.

* `style_weight` (number): Strength of adherence to the new style. Range 0-1, up to 2 decimals.

* `weirdness_constraint` (number): Controls creative deviation. Range 0-1, up to 2 decimals.

* `audio_weight` (number): Balance weight for preserving original audio features. Range 0-1, up to 2 decimals.

* `persona_id` (string): Persona ID to apply to the covered track. How to generate persona\_id, visit [generate-persona](/api-manual/music-series/generate-persona).


# Upload and Extend Audio
Source: https://docs.vidgo.ai/api-manual/music-series/upload-and-extend-audio

api-manual/music-series/upload-and-extend-audio.json POST /api/generate/submit
Extend audio tracks while preserving the original style

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

* This endpoint extends uploaded audio files while preserving the original style
* Upload an existing audio track and specify where to continue from
* The AI generates additional content that seamlessly blends with the original
* Uploaded audio files must not exceed 8 minutes in length (**Note**: For V4\_5ALL model, max 1 minute)

## Parameter Details

* When `default_param_flag: true` (Custom Parameters Mode):

  * Full control over style, title, and other parameters

  * If `instrumental: true`: only `style`, `title`, and `upload_url` are required

  * If `instrumental: false`: `style`, `title`, `prompt` (used as exact lyrics), and `upload_url` are required

  * Character limits vary by model:

    * **V4**: `prompt` 3000 characters, `style` 200 characters, `title` 80 characters

    * **V4\_5 & V4\_5PLUS**: `prompt` 5000 characters, `style` 1000 characters, `title` 100 characters

    * **V4\_5ALL**: `prompt` 5000 characters, `style` 1000 characters, `title` 80 characters

    * **V5**: `prompt` 5000 characters, `style` 1000 characters, `title` 100 characters

* When `default_param_flag: false`:

  * Only `upload_url` is required. If `instrumental: false`, lyrics will be automatically generated

## Usage Notes

* Recommendation: Use `default_param_flag: false` for quick extensions that maintain the original style
* The `continue_at` parameter determines where the extension begins - set it to the point where you want new content to start

## Optional parameters

* `prompt` (string): Description or lyrics for the extended audio. Required when `default_param_flag` is `true` and `instrumental` is `false`. The prompt will be used strictly as lyrics. When `default_param_flag` is `false`, lyrics will be automatically generated based on the prompt.

* `style` (string): Music style specification. Required when `default_param_flag` is `true`.

* `title` (string): Track title for the extended audio. Character limits vary by model (V4: 80, V4\_5 & V4\_5PLUS & V5: 100, V4\_5ALL: 80).

* `negative_tags` (string): Music styles to exclude from generation.

* `vocal_gender` (string): Vocal gender preference. Use `m` for male, `f` for female. Note: This parameter can only increase the probability but cannot guarantee the specified gender.

* `style_weight` (number): Strength of adherence to style. Range 0-1, up to 2 decimals.

* `weirdness_constraint` (number): Controls creative deviation. Range 0-1, up to 2 decimals.

* `audio_weight` (number): Balance weight for audio features. Range 0-1, up to 2 decimals.

* `persona_id` (string): Persona ID to apply to the generated music. How to generate persona\_id, visit [generate-persona](/api-manual/music-series/generate-persona).


# Separate Vocals
Source: https://docs.vidgo.ai/api-manual/music-series/vocal-remover/separate-vocals

api-manual/music-series/vocal-remover/separate-vocals.json POST /api/generate/submit
Separate vocals and instruments from music tracks

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

> [!NOTE]
>
> **Vocal Separation Options:**
>
> * [separate-vocals](/api-manual/music-series/vocal-remover/separate-vocals): Does not support uploading audio, 2-stem split, based on Suno
> * [stem-split](/api-manual/music-series/vocal-remover/stem-split): Does not support uploading audio, 12-stem split, based on Suno
> * [upload-and-separate-vocals](/api-manual/music-series/vocal-remover/upload-and-separate-vocals): Supports uploading audio, 7-stem split

* This endpoint separates vocals and instruments from music tracks
* Useful for remixing, karaoke tracks, or isolating specific instruments

## Parameter Details

* `task_id` (required): Task ID from a completed music generation (Generate Music or Extend Music)
* `audio_id` (required): Specific audio track identifier from the callback data

## Usage Notes

* Separation can only be performed once per audio track
* The callback includes download URLs for each separated track


# Stem Split
Source: https://docs.vidgo.ai/api-manual/music-series/vocal-remover/stem-split

api-manual/music-series/vocal-remover/stem-split.json POST /api/generate/submit
Split music tracks into multiple instrument stems

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

> [!NOTE]
>
> **Vocal Separation Options:**
>
> * [separate-vocals](/api-manual/music-series/vocal-remover/separate-vocals): Does not support uploading audio, 2-stem split, based on Suno
> * [stem-split](/api-manual/music-series/vocal-remover/stem-split): Does not support uploading audio, 12-stem split, based on Suno
> * [upload-and-separate-vocals](/api-manual/music-series/vocal-remover/upload-and-separate-vocals): Supports uploading audio, 7-stem split

* This endpoint splits music tracks into multiple instrument stems
* Creates separate tracks for backing vocals, bass, brass, drums, fx, guitar, keyboard, percussion, strings, synth, vocals, woodwinds, and piano
* Useful for detailed remixing or isolating specific instruments

## Parameter Details

* `task_id` (required): Task ID from a completed music generation (Generate Music or Extend Music)
* `audio_id` (required): Specific audio track identifier from the callback data

## Usage Notes

* Stem split can only be performed once per audio track
* The callback includes download URLs for each separated stem


# Upload and Separate Vocals
Source: https://docs.vidgo.ai/api-manual/music-series/vocal-remover/upload-and-separate-vocals

api-manual/music-series/vocal-remover/upload-and-separate-vocals.json POST /api/generate/submit
Upload audio and separate it into multiple stems

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the response result through the unified [Query Music Detail](/api-manual/music-series/query-music-detail) endpoint.

## Usage Guide

> [!NOTE]
>
> **Vocal Separation Options:**
>
> * [separate-vocals](/api-manual/music-series/vocal-remover/separate-vocals): Does not support uploading audio, 2-stem split, based on Suno
> * [stem-split](/api-manual/music-series/vocal-remover/stem-split): Does not support uploading audio, 12-stem split, based on Suno
> * [upload-and-separate-vocals](/api-manual/music-series/vocal-remover/upload-and-separate-vocals): Supports uploading audio, 7-stem split

* This endpoint uploads audio and separates it into multiple stems
* Creates separate tracks for bass, drums, piano, guitar, vocals, and other
* Useful for remixing or isolating specific instruments from uploaded audio

## Parameter Details

* `audio_url` (required): URL of the audio file to upload and separate
* `title` (optional): Title for the separation task
* `model_name` (optional): Separation model to use. Default: `base`
  * `base`: Standard separation quality
  * `enhanced`: Higher quality separation with better accuracy
  * `instrumental`: Optimized for instrumental tracks
* `output_type` (optional): Type of output stems to generate. Default: `general`
  * `general`: All available stems
  * `bass`: Bass track only
  * `drums`: Drums track only
  * `other`: Other instruments
  * `piano`: Piano track only
  * `guitar`: Guitar track only
  * `vocals`: Vocals track only


# xAI TTS 1
Source: https://docs.vidgo.ai/api-manual/music-series/xai-tts-v1

api-manual/music-series/xai-tts-v1.json POST /api/generate/submit
Convert text into speech with xAI TTS, choosing from five voices and multiple languages for narration, dialogue, and spoken content.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Convert text into speech with xAI TTS, choosing from five voices and multiple languages for narration, dialogue, and spoken content.

## Models

* **xai/tts/v1** - Convert text into speech.

## Features

* Convert text into speech.
* Select a voice with `voice` and specify the spoken language with `language_code`.
* Direct pauses, laughter, sighs, whispers, and slower phrases through tags in `text`.

## Output Options

### Output Format

* `output_format`: Optional. Supply an object with the fields below.

* `output_format.codec`: Optional. Accepted values: `mp3`, `wav`, `pcm`, `mulaw`, `alaw`. Default: `mp3`.

* `output_format.sample_rate`: Optional. Sample rate in Hz. Accepted values: `8000`, `16000`, `22050`, `24000`, `44100`, `48000`. Default: `24000`.

* `output_format.bit_rate`: Optional. MP3 bitrate in bits per second. Accepted values: `32000`, `64000`, `96000`, `128000`, `192000`, `null`. MP3 default: `128000`.

* For `mp3`, use `output_format.bit_rate` to select the bitrate. For `wav`, `pcm`, `mulaw`, and `alaw`, omit `bit_rate` or set it to `null`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Text

* `text`: Required. Supply the text to speak. Length: **1-15,000 characters**.
* Enter spoken text after trimming leading and trailing whitespace. Tags, punctuation, and spaces within the text count toward the length.
* Add `[laugh]`, `[pause]`, or `[sigh]` where the sound belongs. Use `<whisper>text</whisper>` for whispered passages and `<slow>text</slow>` for slower delivery.

### Voice

* `voice`: Optional. Select the voice. Accepted values: `eve`, `ara`, `rex`, `sal`, `leo`. Default: `eve`.

### Language Code

* `language_code`: Optional. String. Default: `auto`.

> [accordion] Supported language code values
>
> * `auto`
> * `en`
> * `ar-EG`
> * `ar-SA`
> * `ar-AE`
> * `bn`
> * `zh`
> * `fr`
> * `de`
> * `hi`
> * `id`
> * `it`
> * `ja`
> * `ko`
> * `pt-BR`
> * `pt-PT`
> * `ru`
> * `es-MX`
> * `es-ES`
> * `tr`
> * `vi`

## Output Files

* Read every item in the returned `files` array.
* Generated audio uses `file_type: "audio"`.
* Use `file_url` to download the generated speech.

## Request Example

```json theme={null}
{
  "model": "xai/tts/v1",
  "input": {
    "text": "Welcome to our story. [pause] <whisper>Listen closely.</whisper> [laugh] Let us begin.",
    "voice": "eve",
    "language_code": "auto",
    "output_format": {
      "codec": "mp3",
      "sample_rate": 24000,
      "bit_rate": 128000
    }
  }
}
```


# OmniHuman 2.0 Audio to Video
Source: https://docs.vidgo.ai/api-manual/video-series/bytedance-omnihuman-2-0-audio-to-video

api-manual/video-series/bytedance-omnihuman-2-0-audio-to-video.json POST /api/generate/submit
The OmniHuman 2.0 Audio to Video API animates one character image with a 4–300 second WAV or MP3 file into a 720p lip-synced video, with optional prompts.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

OmniHuman 2.0 Audio to Video takes one character image and one audio file and generates a 720p video of that character speaking, singing or performing, with lip movement, expression and gestures following the audio. It accepts 4–300 seconds of audio, so it works for short social clips as well as full explainer narrations. OmniHuman 2.0 is ByteDance's audio-driven video model, offered upstream through BytePlus.

## Models

* **bytedance/omnihuman-2.0/audio-to-video** — Audio-to-video with OmniHuman 2.0: one image plus one WAV or MP3 file, 720p output, length follows the audio.

## Features

* Audio-driven talking and singing videos from a single image.
* Lip sync, expression and gestures driven by the audio.
* Audio from 4 seconds up to 5 minutes; the video is as long as the audio.
* Works with people, animals, cartoon, 3D and stylized characters, and multi-person images.
* Optional prompt for camera movement, actions and added scene elements.

## Required parameters

* `image_urls` (string\[]): Exactly one public HTTP(S) image URL of the character. The framing of the video follows this image.
* `audio_url` (string): Public HTTP(S) URL of a WAV or MP3 file, at most 15 MB and 4–300 seconds long.

## Optional parameters

* `prompt` (string): Camera movement, actions or extra scene elements. Lip movement and expression still come from the audio.

## Duration

There is no `duration` field. The output length follows the audio, so a 30-second voice-over gives a video of about 30 seconds. Trim the audio to change the length.

## Output

The finished task returns one MP4 video in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload. The video is 720p.

## Examples

* **Talking presenter from a headshot**: one headshot and an MP3 voice-over, no prompt.
* **Singing character with a camera prompt and callback**: a cartoon cat, a WAV song, a prompt for movement and camera, and a `callback_url`.

## Notes

* Use one clear image with a visible face and clean audio without long silences or heavy background noise for the most accurate lip sync.
* Audio that is not WAV or MP3, larger than 15 MB, shorter than 4 seconds or longer than 300 seconds returns 400. `duration`, `resolution`, `aspect_ratio`, `seed` and any other field not listed above also return 400.
* `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are billed per second of generated video. At submission, credits for the full audio length are reserved. When the task succeeds, the charge is settled on the actual output length and any difference is refunded; the final charge never exceeds the reserved amount. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Seedance 1.5 Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/bytedance-seedance-v1-5-pro-image-to-video

api-manual/video-series/bytedance-seedance-v1-5-pro-image-to-video.json POST /api/generate/submit
Animate reference images with Seedance 1.5 Pro, with optional generated audio and fixed-camera control at up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate reference images with Seedance 1.5 Pro, with optional generated audio and fixed-camera control at up to 1080p.

## Models

* **bytedance/seedance-v1.5-pro/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Accepts 1-2 images in `image_urls`.
* Choose from `480p`, `720p`, `1080p` output resolutions.

## Duration Options

* **4, 8, 12 seconds** - Set `duration` as an integer. Required.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **3-2,500 characters**.
* Surrounding whitespace is trimmed before counting Unicode characters.

### Image URLs

* `image_urls`: Required. Supply 1–2 publicly accessible HTTP(S) image URLs. The first image is the first frame; the optional second image is the last frame.
* Use JPEG, PNG or WebP images, up to **10 MB per image**.

### Aspect Ratio

* `aspect_ratio`: Required. Accepted values: `1:1`, `21:9`, `4:3`, `3:4`, `16:9`, `9:16`.

### Resolution

* `resolution`: Optional. Accepted values: `480p`, `720p`, `1080p`. Vidgo defaults to `720p` when omitted.

### Fixed Lens

* `fixed_lens`: Optional. Control fixed-camera generation. Use `true` or `false`.

### Generate Audio

* `generate_audio`: Optional. Set `true` to generate synchronized audio or `false` for a silent video. Vidgo defaults to `true` when omitted.


# Seedance 1.5 Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/bytedance-seedance-v1-5-pro-text-to-video

api-manual/video-series/bytedance-seedance-v1-5-pro-text-to-video.json POST /api/generate/submit
Create video scenes with Seedance 1.5 Pro, with optional generated audio and a fixed-camera setting for controlled compositions.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create video scenes with Seedance 1.5 Pro, with optional generated audio and a fixed-camera setting for controlled compositions.

## Models

* **bytedance/seedance-v1.5-pro/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `480p`, `720p`, `1080p` output resolutions.

## Duration Options

* **4, 8, 12 seconds** - Set `duration` as an integer. Required.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **3-2,500 characters**.
* Surrounding whitespace is trimmed before counting Unicode characters.

### Aspect Ratio

* `aspect_ratio`: Required. Accepted values: `1:1`, `21:9`, `4:3`, `3:4`, `16:9`, `9:16`.

### Resolution

* `resolution`: Optional. Accepted values: `480p`, `720p`, `1080p`. Vidgo defaults to `720p` when omitted.

### Fixed Lens

* `fixed_lens`: Optional. Control fixed-camera generation. Use `true` or `false`.

### Generate Audio

* `generate_audio`: Optional. Set `true` to generate synchronized audio or `false` for a silent video. Vidgo defaults to `true` when omitted.


# Seedance 1.0 Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/bytedance-seedance-v1-pro-image-to-video

api-manual/video-series/bytedance-seedance-v1-pro-image-to-video.json POST /api/generate/submit
Turn a still image into a five- or ten-second video with Seedance 1.0 Pro, using text to guide the scene and motion.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Turn a still image into a five- or ten-second video with Seedance 1.0 Pro, using text to guide the scene and motion.

## Models

* **bytedance/seedance/v1/pro/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Accepts exactly 1 image in `image_urls`.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **5, 10 seconds** - Set `duration` as a numeric integer. Optional. Default: **5 seconds**, only when omitted. Strings, booleans, null and fractional values are rejected. A numeric value such as `5.0` is equivalent to `5`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials or whitespace. Defaults apply only when optional input fields are omitted; explicit null and empty values are rejected.

### Prompt

* `prompt`: Required string. Describe the scene and desired motion. After trimming surrounding whitespace, length must be **1-10,000 Unicode characters**. Non-string values are rejected.
* The value must not be blank when supplied.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL. Base64, a bare string instead of an array, and empty arrays are rejected.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`, only when omitted. Aliases, different casing, surrounding whitespace and numeric values are rejected.

## Billing Factors

Billing depends on output resolution (`resolution`) and video duration (`duration`). Text-to-video and image-to-video use the same rates.

Credits are deducted on submission and refunded if the task fails.


# Seedance 1.0 Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/bytedance-seedance-v1-pro-text-to-video

api-manual/video-series/bytedance-seedance-v1-pro-text-to-video.json POST /api/generate/submit
Create five- or ten-second video scenes from written descriptions with Seedance 1.0 Pro, with output up to 1080p.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create five- or ten-second video scenes from written descriptions with Seedance 1.0 Pro, with output up to 1080p.

## Models

* **bytedance/seedance/v1/pro/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Choose from `720p`, `1080p` output resolutions.

## Duration Options

* **5, 10 seconds** - Set `duration` as a numeric integer. Optional. Default: **5 seconds**, only when omitted. Strings, booleans, null and fractional values are rejected. A numeric value such as `5.0` is equivalent to `5`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Defaults apply only when optional input fields are omitted; explicit null and empty values are rejected.

### Prompt

* `prompt`: Required string. Describe the scene and desired motion. After trimming surrounding whitespace, length must be **1-10,000 Unicode characters**. Non-string values are rejected.
* The value must not be blank when supplied.

### Resolution

* `resolution`: Optional. Accepted values: `720p`, `1080p`. Default: `720p`, only when omitted. Aliases, different casing, surrounding whitespace and numeric values are rejected.

## Billing Factors

Billing depends on output resolution (`resolution`) and video duration (`duration`). Text-to-video and image-to-video use the same rates.

Credits are deducted on submission and refunded if the task fails.


# MiniMax H3 Max Turbo Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-h3-max-turbo-image-to-video

api-manual/video-series/minimax-h3-max-turbo-image-to-video.json POST /api/generate/submit
The MiniMax H3 Max Turbo Image to Video API animates a start frame and optional end frame, given as URLs or Base64, into a 5–15 second 480p–1080p video.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `enable_safety_checker`: boolean, defaults to `true`. Set `false` to disable it.
* `callback_url`: optional webhook URL at the request root. Query task status if you do not use a callback.

## Usage Guide

Use `minimax/h3-max-turbo/image-to-video`. Provide `image_urls` in order: the starting frame first, then an optional ending frame. The output follows the starting image; omit `aspect_ratio`.

Set `model` and optional `callback_url` at the request root. Place generation parameters inside `input`.

## Parameter Details

* `prompt`: required non-empty string, up to 50,000 characters.
* `duration`: integer from 5 through 15 seconds; defaults to 5. Fractional durations are rejected.
* `resolution`: `480p`, `768p` or `1080p`; defaults to `768p`. Uppercase `P` is also accepted.
* `image_urls`: required array of 1-2 images, in starting-frame and ending-frame order. Public HTTP(S) URLs and Base64 image data are supported.


# MiniMax H3 Max Turbo Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-h3-max-turbo-text-to-video

api-manual/video-series/minimax-h3-max-turbo-text-to-video.json POST /api/generate/submit
The MiniMax H3 Max Turbo Text to Video API is the Turbo variant of H3 Max, turning a prompt into a 5–15 second 480p–1080p video in six aspect ratios.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Optional parameters

* `enable_safety_checker`: boolean, defaults to `true`. Set `false` to disable it.
* `callback_url`: optional webhook URL at the request root. Query task status if you do not use a callback.

## Usage Guide

Use `minimax/h3-max-turbo/text-to-video`. Provide a prompt without media. Select the text endpoint for prompt-only generation.

Set `model` and optional `callback_url` at the request root. Place generation parameters inside `input`.

## Parameter Details

* `prompt`: required non-empty string, up to 50,000 characters.
* `duration`: integer from 5 through 15 seconds; defaults to 5. Fractional durations are rejected.
* `resolution`: `480p`, `768p` or `1080p`; defaults to `768p`. Uppercase `P` is also accepted.
* `aspect_ratio`: `21:9`, `16:9`, `4:3`, `1:1`, `3:4` or `9:16`; defaults to `16:9`.


# Hailuo 02 Pro Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-hailuo-02-pro-image-to-video

api-manual/video-series/minimax-hailuo-02-pro-image-to-video.json POST /api/generate/submit
Animate still images with Hailuo 02 Pro, with text guidance, prompt refinement, and an optional closing frame to shape the shot.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate still images with Hailuo 02 Pro, with text guidance, prompt refinement, and an optional closing frame to shape the shot.

## Models

* **minimax/hailuo-02/pro/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Accepts exactly 1 image in `image_urls`.
* Optionally supply an ending frame with `end_image_url`.

## Output Options

* `resolution`: Optional. Accepted values: `512P`, `768P`. Default: `768P`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. No maximum prompt length is specified.
* Whitespace is trimmed before validation; the result must not be blank.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.

### End Image URL

* `end_image_url`: Optional. Provide the ending frame URL. Use a publicly accessible HTTP(S) URL. Requires a starting image in `image_urls`. The ending frame is supported at both resolutions.

### Prompt Optimizer

* `prompt_optimizer`: Optional. Control automatic prompt optimization. Use `true` or `false`.

## Billing Factors

Billed per generation using the Pro model tier. Both supported resolutions use the same rate.

Credits are deducted on submission and refunded if the task fails.


# Hailuo 02 Pro Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-hailuo-02-pro-text-to-video

api-manual/video-series/minimax-hailuo-02-pro-text-to-video.json POST /api/generate/submit
Create short scenes from text with Hailuo 02 Pro, with optional prompt refinement to develop the visual direction of a clip.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create short scenes from text with Hailuo 02 Pro, with optional prompt refinement to develop the visual direction of a clip.

## Models

* **minimax/hailuo-02/pro/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.

## Output Options

* `resolution`: Optional. Accepted values: `512P`, `768P`. Default: `768P`.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. No maximum prompt length is specified.
* Whitespace is trimmed before validation; the result must not be blank.

### Prompt Optimizer

* `prompt_optimizer`: Optional. Control automatic prompt optimization. Use `true` or `false`.

## Billing Factors

Billed per generation using the Pro model tier. Both supported resolutions use the same rate.

Credits are deducted on submission and refunded if the task fails.


# Hailuo 02 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-hailuo-02-standard-image-to-video

api-manual/video-series/minimax-hailuo-02-standard-image-to-video.json POST /api/generate/submit
Animate still images with Hailuo 02 Standard, with text guidance, prompt refinement, and an optional closing frame to shape the shot.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Animate still images with Hailuo 02 Standard, with text guidance, prompt refinement, and an optional closing frame to shape the shot.

## Models

* **minimax/hailuo-02/standard/image-to-video** - Generate video from image input.

## Features

* Generate video from image input.
* Accepts exactly 1 image in `image_urls`.
* Optionally supply an ending frame with `end_image_url`.
* Accepts `512P` or `768P` resolution; defaults to `768P`.

## Duration Options

* **6, 10 seconds** - Set `duration` as an integer. Optional. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted. Input URLs must use HTTP(S), include a hostname, and contain no embedded credentials.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-1,500 characters**.
* Whitespace is trimmed before validation; the result must not be blank.

### Image URLs

* `image_urls`: Required. Provide publicly accessible image URLs. Supply exactly 1 item. Each item must be an HTTP(S) URL.

### End Image URL

* `end_image_url`: Optional. Provide the ending frame URL. Use a publicly accessible HTTP(S) URL. Requires a starting image in `image_urls`. Standard requires `768P` when an ending frame is supplied.

### Resolution

* `resolution`: Optional. Accepted values: `512P`, `768P`. Default: `768P`.

### Prompt Optimizer

* `prompt_optimizer`: Optional. Control automatic prompt optimization. Use `true` or `false`.

## Billing Factors

Billing depends on output resolution (`resolution`) and video duration (`duration`). Usage is charged per generated second using the Standard model tier.

Credits are deducted on submission and refunded if the task fails.


# Hailuo 02 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/minimax-hailuo-02-standard-text-to-video

api-manual/video-series/minimax-hailuo-02-standard-text-to-video.json POST /api/generate/submit
Create short scenes from text with Hailuo 02 Standard, with optional prompt refinement to develop the visual direction of a clip.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Create short scenes from text with Hailuo 02 Standard, with optional prompt refinement to develop the visual direction of a clip.

## Models

* **minimax/hailuo-02/standard/text-to-video** - Generate video scenes from text prompts.

## Features

* Generate video scenes from text prompts.
* Accepts `512P` or `768P` resolution; defaults to `768P`.

## Duration Options

* **6, 10 seconds** - Set `duration` as an integer. Optional. Default: **6 seconds**.

## Advanced Parameters

Set `model` and optional `callback_url` at the request root. Place all workflow parameters, including output options, inside `input`.

Send only the input fields supported by this workflow. Unsupported `input` fields are rejected before credits are deducted.

### Prompt

* `prompt`: Required. Describe the scene and desired motion. Length: **1-1,500 characters**.
* Whitespace is trimmed before validation; the result must not be blank.

### Resolution

* `resolution`: Optional. Accepted values: `512P`, `768P`. Default: `768P`.

### Prompt Optimizer

* `prompt_optimizer`: Optional. Control automatic prompt optimization. Use `true` or `false`.

## Billing Factors

Billing depends on output resolution (`resolution`) and video duration (`duration`). Usage is charged per generated second using the Standard model tier.

Credits are deducted on submission and refunded if the task fails.


# PixVerse V6 Max Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/pixverse-v6-max-image-to-video

api-manual/video-series/pixverse-v6-max-image-to-video.json POST /api/generate/submit
The PixVerse V6 Max Image to Video API animates one JPEG, PNG or WebP image into a 1–15 second video at 360p to 1080p, keeping the image's framing.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

PixVerse V6 Max Image to Video animates one JPEG, PNG or WebP image into a 1–15 second video at 360p to 1080p. The image is the start frame and keeps its framing, so the output follows the image's aspect ratio. Add `audio: true` for a native soundtrack. It suits product animations from a packshot, social clips and music-video snippets.

PixVerse V6 Max is the Max tier of PixVerse V6, added to PixVerse's API for text-to-video and image-to-video in October 2026. Vidgo exposes two model IDs: [Text to Video](/api-manual/video-series/pixverse-v6-max-text-to-video) and [Image to Video](/api-manual/video-series/pixverse-v6-max-image-to-video).

## Models

* **pixverse/v6-max/image-to-video** — Image-to-video with PixVerse V6 Max from one image, 1–15 seconds, 360p to 1080p, optional audio.

## Features

* Image-to-video from a single start image.
* Optional prompt to direct motion and camera movement.
* Durations from 1 to 15 seconds in one-second steps.
* Four resolutions: `360p`, `540p`, `720p` and `1080p`.
* Optional native audio track with `audio: true`.
* Aspect ratio follows the input image.

## Required parameters

* `image_urls` (string\[]): Exactly one public HTTP(S) image URL (JPEG, PNG or WebP). If the URL has a file extension, it must be `.jpg`, `.jpeg`, `.png` or `.webp`.

## Optional parameters

* `prompt` (string): What moves, how the camera moves and, with `audio` on, the sound. Must not be blank if sent; up to 2,048 characters.
* `duration` (integer): Output length, `1`–`15` whole seconds. Default `5`.
* `resolution` (string): `360p`, `540p`, `720p` or `1080p`. Default `540p`.
* `audio` (boolean): Add a native audio track. Default `false`.

## Output

The finished task returns one MP4 video in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **5-second 720p product animation**: a packshot with a short motion prompt, silent.
* **8-second 1080p clip with audio and callback**: a concert photo with `audio: true` and a `callback_url`.

## Notes

* `aspect_ratio` is not accepted on this ID (the image sets it) and returns 400, as does any field not listed above. For prompt-only generation with a chosen aspect ratio, use [PixVerse V6 Max Text to Video](/api-manual/video-series/pixverse-v6-max-text-to-video).
* The prompt limit here is 2,048 characters, lower than the 5,000 allowed in text-to-video.
* PixVerse V6 Max on Vidgo takes text or exactly one image. First/last frame, video input, extend, reference-to-video and `seed` are not available.
* A `null` value for `duration`, `resolution` or `audio` counts as the default.
* `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution`, `audio` and `duration` (per output second); turning on `audio` raises the rate. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# PixVerse V6 Max Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/pixverse-v6-max-text-to-video

api-manual/video-series/pixverse-v6-max-text-to-video.json POST /api/generate/submit
The PixVerse V6 Max Text to Video API generates 1–15 second videos from a prompt at 360p to 1080p, with eight aspect ratios and an optional native audio track.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

PixVerse V6 Max Text to Video turns a prompt into a 1–15 second video at 360p to 1080p, with an optional native audio track for dialogue, effects and music. Low resolutions make quick, inexpensive drafts; 1080p with audio suits finished social ads, short-drama scenes and cinematic `21:9` teasers.

PixVerse V6 Max is the Max tier of PixVerse V6, added to PixVerse's API for text-to-video and image-to-video in October 2026. Vidgo exposes two model IDs: [Text to Video](/api-manual/video-series/pixverse-v6-max-text-to-video) and [Image to Video](/api-manual/video-series/pixverse-v6-max-image-to-video).

## Models

* **pixverse/v6-max/text-to-video** — Text-to-video with PixVerse V6 Max, 1–15 seconds, 360p to 1080p, optional audio.

## Features

* Text-to-video from a single prompt.
* Durations from 1 to 15 seconds in one-second steps.
* Four resolutions: `360p`, `540p`, `720p` and `1080p`.
* Optional native audio track with `audio: true`.
* Eight aspect ratios, including `21:9`, `2:3` and `3:2`.

## Required parameters

* `prompt` (string): The scene, the action and the camera. With `audio` on, put dialogue in quotes and name the speaker. Must not be blank; up to 5,000 characters.

## Optional parameters

* `duration` (integer): Output length, `1`–`15` whole seconds. Default `5`.
* `resolution` (string): `360p`, `540p`, `720p` or `1080p`. Default `540p`.
* `audio` (boolean): Add a native audio track. Default `false`.
* `aspect_ratio` (string): `16:9`, `9:16`, `4:3`, `3:4`, `1:1`, `2:3`, `3:2` or `21:9`. When omitted, the upstream default (16:9) applies.

## Output

The finished task returns one MP4 video in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **5-second 540p silent draft**: a quick `16:9` preview at the default resolution, `audio: false`.
* **10-second vertical 1080p clip with dialogue**: quoted dialogue plus ambience in the prompt, `9:16`, `audio: true` and a `callback_url`.

## Notes

* This ID accepts text only; `image_urls` returns 400. To animate an image, use [PixVerse V6 Max Image to Video](/api-manual/video-series/pixverse-v6-max-image-to-video).
* PixVerse V6 Max on Vidgo takes text or exactly one image. First/last frame, video input, extend, reference-to-video and `seed` are not available.
* A `null` value for `duration`, `resolution` or `audio` counts as the default.
* `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution`, `audio` and `duration` (per output second); turning on `audio` raises the rate. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Runway Gen-4.5
Source: https://docs.vidgo.ai/api-manual/video-series/runway-gen-4-5

api-manual/video-series/runway-gen-4-5.json POST /api/generate/submit
Video generation with optional reference image, 5s and 10s durations

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

`runway-gen-4.5` is exposed through the unified submit endpoint with text-to-video and optional single-image guidance.

## Models

* **runway-gen-4.5** - Text-to-video generation

## Required parameters

* **prompt**: Text prompt for generation

## Optional parameters

* **duration**: `5` or `10`. Default is `5`
* **aspect\_ratio**: `16:9`, `9:16`, `4:3`, `3:4`, `1:1`, or `21:9`. Default is `16:9`
* **image\_urls**: Optional image URL list. At most `1` image
* **seed**: Optional integer seed

## Notes

* `image_urls` is optional. If provided, only one image URL is allowed


# Seedance 1.5 Pro
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-1-5-pro

api-manual/video-series/seedance-1-5-pro.json POST /api/generate/submit
High-quality text-to-video and image-to-video generation with Seedance 1.5 Pro

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate high-quality videos from text prompts or reference images using Seedance 1.5 Pro by ByteDance. Control aspect ratio, resolution, duration, and optional audio generation.

## Models

* **seedance-1.5-pro** - Text-to-video and image-to-video generation with enhanced fidelity

## Features

* **Flexible Modes**: Supports text-to-video and image-to-video generation
* **Aspect Ratio Control**: Flexible output aspect ratios
* **Resolution Control**: Choose the target output resolution
* **Duration Control**: Set duration in seconds
* **Fixed Lens**: Lock the camera lens for steady framing
* **Audio Generation**: Optionally generate an audio track

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [bytedance/seedance-v1.5-pro/text-to-video](/api-manual/video-series/bytedance-seedance-v1-5-pro-text-to-video)
* [bytedance/seedance-v1.5-pro/image-to-video](/api-manual/video-series/bytedance-seedance-v1-5-pro-image-to-video)


# Seedance 1.0 Pro
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-1.0-pro

api-manual/video-series/seedance-1.0-pro.json POST /api/generate/submit
High-quality text-to-video and image-to-video generation with Seedance 1.0 Pro

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Generate high-quality videos from text prompts or reference images using Seedance 1.0 Pro by ByteDance. Create smooth, professional animations with flexible duration and resolution options.

## Models

* **seedance-1.0-pro** - Text-to-video and image-to-video generation with 3x faster rendering

## Features

* **Flexible Modes**: Supports text-to-video and image-to-video generation
* **Multiple Resolutions**: 720p and 1080p support
* **Duration Options**: 5 or 10 seconds
* **Fast Rendering**: 3x faster than standard Pro version

## Workflow IDs

Existing IDs remain available. New integrations can select a dedicated workflow:

* [bytedance/seedance/v1/pro/text-to-video](/api-manual/video-series/bytedance-seedance-v1-pro-text-to-video)
* [bytedance/seedance/v1/pro/image-to-video](/api-manual/video-series/bytedance-seedance-v1-pro-image-to-video)


# Seedance 2
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2

api-manual/video-series/seedance-2.json POST /api/generate/submit
Text-to-video, first and last frame, and multimodal reference generation with Seedance 2 and Seedance 2 Fast

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Seedance 2 is ByteDance's multimodal video family for API-based generation. On Vidgo API, you can choose between **seedance-2** for higher fidelity and **seedance-2-fast** for lower-latency generation, while keeping a single submit and status API.

## Models

* **seedance-2**: Higher-quality video generation with multimodal references and native audio support
* **seedance-2-fast**: Faster generation path with the same submit schema and lower per-second pricing

## Supported Generation Modes

* **Text to Video**: Generate directly from a prompt
* **First and Last Frame**: Guide motion between one or two frame anchors via `image_urls`
* **Multimodal Reference**: Add reference images, videos, and audio through `reference_image_urls`, `reference_video_urls`, and `reference_audio_urls`

## Input Rules

* `resolution`: `seedance-2` supports `480p`, `720p`, `1080p`, and `4k`; `seedance-2-fast` supports `480p` and `720p`
* `duration` accepts integer values from `4` to `15`
* `image_urls` supports up to `2` items
* `image_urls` is mutually exclusive with all `reference_*_urls` fields
* `reference_image_urls` supports up to `9` items; `reference_video_urls` and `reference_audio_urls` each support up to `3` items
* `reference_audio_urls` requires at least one `reference_image_urls` or `reference_video_urls` item
* Total reference inputs across image, video, and audio fields can contain at most `12` files
* `aspect_ratio` supports `auto`, `1:1`, `21:9`, `4:3`, `3:4`, `16:9`, and `9:16`
* `generate_audio`, `return_last_frame`, and `web_search` are optional boolean controls

## Seedance 2 vs Seedance 1.5 Pro vs Seedance 1.0 Pro

* **Seedance 2** adds multimodal reference image, video, and audio guidance with first and last frame control in one API shape.
* **Seedance 1.5 Pro** focuses on high-quality text-to-video and image-to-video generation with optional audio but without the same multimodal reference set.
* **Seedance 1.0 Pro** remains a strong option for earlier Seedance use cases, but Seedance 2 expands creative control and API flexibility for newer production pipelines.


# Seedance 2.0 Fast Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-fast-image-to-video

api-manual/video-series/seedance-2-0-fast-image-to-video.json POST /api/generate/submit
Seedance 2.0 Fast image to video model with multimodal generation and optional audio

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0-fast/image-to-video** - Seedance 2.0 Fast image to video generation

## Features

* **Image To Video**: Control the starting frame and optional ending frame with up to two images.
* **Optional Audio**: Request generated audio together with the video.

## Variant Notes

* This public model ID uses the `seedance-2-fast` provider variant.
* Supported resolutions are `480p` and `720p`.

## Credits Reference

* Usage is calculated from output duration and selected resolution.
* Check the Vidgo dashboard for the current model rate before submitting production traffic.

## Notes

* `prompt` is required and accepts up to 20,000 characters.
* `duration` supports integer values from `4` to `15`.
* `image_urls` accepts one starting frame and one optional ending frame.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* `image_urls` is required.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.0 Fast Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-fast-reference-to-video

api-manual/video-series/seedance-2-0-fast-reference-to-video.json POST /api/generate/submit
Seedance 2.0 Fast reference to video model with multimodal generation and optional audio

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0-fast/reference-to-video** - Seedance 2.0 Fast reference to video generation

## Features

* **Reference To Video**: Guide generation with reference images, videos, and optional audio.
* **Optional Audio**: Request generated audio together with the video.

## Variant Notes

* This public model ID uses the `seedance-2-fast` provider variant.
* Supported resolutions are `480p` and `720p`.

## Credits Reference

* Usage is calculated from selected resolution and billed duration. Each reference-video duration, rounded down separately, is added to output duration.
* Check the Vidgo dashboard for the current model rate before submitting production traffic.

## Notes

* `prompt` is required and accepts up to 20,000 characters.
* `duration` supports integer values from `4` to `15`.
* Reference input supports up to 9 images, 3 videos, and 3 audio files, with at most 12 files in total. Reference audio requires at least one reference image or reference video.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* Provide at least one reference image or reference video. Reference audio cannot be used alone.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.0 Fast Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-fast-text-to-video

api-manual/video-series/seedance-2-0-fast-text-to-video.json POST /api/generate/submit
Seedance 2.0 Fast text to video model with multimodal generation and optional audio

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0-fast/text-to-video** - Seedance 2.0 Fast text to video generation

## Features

* **Text To Video**: Generate video directly from prompt text.
* **Optional Audio**: Request generated audio together with the video.

## Variant Notes

* This public model ID uses the `seedance-2-fast` provider variant.
* Supported resolutions are `480p` and `720p`.

## Credits Reference

* Usage is calculated from output duration and selected resolution.
* Check the Vidgo dashboard for the current model rate before submitting production traffic.

## Notes

* `prompt` is required and accepts up to 20,000 characters.
* `duration` supports integer values from `4` to `15`.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.0 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-image-to-video

api-manual/video-series/seedance-2-0-image-to-video.json POST /api/generate/submit
Seedance 2.0 image to video model with multimodal generation and optional audio

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0/image-to-video** - Seedance 2.0 image to video generation

## Features

* **Image To Video**: Control the starting frame and optional ending frame with up to two images.
* **Optional Audio**: Request generated audio together with the video.

## Variant Notes

* This public model ID uses the `seedance-2` provider variant.
* Supported resolutions are `480p`, `720p`, `1080p`, and `4k`.

## Credits Reference

* Usage is calculated from output duration and selected resolution.
* Check the Vidgo dashboard for the current model rate before submitting production traffic.

## Notes

* `prompt` is required and accepts up to 20,000 characters.
* `duration` supports integer values from `4` to `15`.
* `image_urls` accepts one starting frame and one optional ending frame.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* `image_urls` is required.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.0 Mini Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-mini-image-to-video

api-manual/video-series/seedance-2-0-mini-image-to-video.json POST /api/generate/submit
Seedance 2.0 Mini image to video model optimized for fast video generation

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0-mini/image-to-video** - Fast Seedance 2.0 Mini image to video generation

## Features

* **Image To Video**: Use up to two images as the starting and optional ending frames.
* **Optional Audio**: Request generated audio together with the video.

## Notes

* `prompt` and `duration` are required.
* `duration` supports integer values from `4` to `15`.
* `resolution` supports `480p` and `720p` and defaults to `720p`.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* `image_urls` accepts one starting frame and one optional ending frame.
* `image_urls` is required.
* Usage is billed by generated video duration and selected resolution.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.0 Mini Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-mini-reference-to-video

api-manual/video-series/seedance-2-0-mini-reference-to-video.json POST /api/generate/submit
Seedance 2.0 Mini reference to video model optimized for fast video generation

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0-mini/reference-to-video** - Fast Seedance 2.0 Mini reference to video generation

## Features

* **Reference To Video**: Guide generation with reference images, videos, and optional audio.
* **Optional Audio**: Request generated audio together with the video.

## Notes

* `prompt` and `duration` are required.
* `duration` supports integer values from `4` to `15`.
* `resolution` supports `480p` and `720p` and defaults to `720p`.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* Reference input supports up to 9 images, 3 videos, and 3 audio files, with at most 12 files in total. Reference audio requires at least one reference image or reference video.
* Provide at least one reference image or reference video. Reference audio cannot be used alone.
* Usage is billed by resolution and billed duration. Each reference-video duration, rounded down separately, is added to output duration.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.0 Mini Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-mini-text-to-video

api-manual/video-series/seedance-2-0-mini-text-to-video.json POST /api/generate/submit
Seedance 2.0 Mini text to video model optimized for fast video generation

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0-mini/text-to-video** - Fast Seedance 2.0 Mini text to video generation

## Features

* **Text To Video**: Generate video directly from prompt text.
* **Optional Audio**: Request generated audio together with the video.

## Notes

* `prompt` and `duration` are required.
* `duration` supports integer values from `4` to `15`.
* `resolution` supports `480p` and `720p` and defaults to `720p`.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* Usage is billed by generated video duration and selected resolution.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.0 Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-reference-to-video

api-manual/video-series/seedance-2-0-reference-to-video.json POST /api/generate/submit
Seedance 2.0 reference to video model with multimodal generation and optional audio

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0/reference-to-video** - Seedance 2.0 reference to video generation

## Features

* **Reference To Video**: Guide generation with reference images, videos, and optional audio.
* **Optional Audio**: Request generated audio together with the video.

## Variant Notes

* This public model ID uses the `seedance-2` provider variant.
* Supported resolutions are `480p`, `720p`, `1080p`, and `4k`.

## Credits Reference

* Usage is calculated from selected resolution and billed duration. Each reference-video duration, rounded down separately, is added to output duration.
* Check the Vidgo dashboard for the current model rate before submitting production traffic.

## Notes

* `prompt` is required and accepts up to 20,000 characters.
* `duration` supports integer values from `4` to `15`.
* Reference input supports up to 9 images, 3 videos, and 3 audio files, with at most 12 files in total. Reference audio requires at least one reference image or reference video.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* Provide at least one reference image or reference video. Reference audio cannot be used alone.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.0 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-0-text-to-video

api-manual/video-series/seedance-2-0-text-to-video.json POST /api/generate/submit
Seedance 2.0 text to video model with multimodal generation and optional audio

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.0/text-to-video** - Seedance 2.0 text to video generation

## Features

* **Text To Video**: Generate video directly from prompt text.
* **Optional Audio**: Request generated audio together with the video.

## Variant Notes

* This public model ID uses the `seedance-2` provider variant.
* Supported resolutions are `480p`, `720p`, `1080p`, and `4k`.

## Credits Reference

* Usage is calculated from output duration and selected resolution.
* Check the Vidgo dashboard for the current model rate before submitting production traffic.

## Notes

* `prompt` is required and accepts up to 20,000 characters.
* `duration` supports integer values from `4` to `15`.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* All remote URLs must be publicly accessible and directly downloadable.


# Seedance 2.5 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-5-image-to-video

api-manual/video-series/seedance-2-5-image-to-video.json POST /api/generate/submit
Seedance 2.5 image to video model with native audio and precise motion control

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.5/image-to-video** - Seedance 2.5 image to video generation

## Input modes

* This model uses one starting frame and an optional ending frame through `image_urls`.
* `image_urls` is required.

## Notes

* `prompt`, `duration`, and `resolution` are required.
* `duration` accepts integer values from 4 to 30 seconds.
* `resolution` accepts `480p`, `720p`, or `1080p`.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* `image_urls` accepts one starting frame and one optional ending frame.
* All media URLs must be publicly accessible and directly downloadable.
* Usage is billed by output duration and selected resolution.


# Seedance 2.5 Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-5-reference-to-video

api-manual/video-series/seedance-2-5-reference-to-video.json POST /api/generate/submit
Seedance 2.5 reference to video model with native audio and precise motion control

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.5/reference-to-video** - Seedance 2.5 reference to video generation

## Input modes

* This model uses reference images, videos, and optional audio for multimodal guidance.
* Provide at least one reference image or reference video. Reference audio cannot be used alone.

## Notes

* `prompt`, `duration`, and `resolution` are required.
* `duration` accepts integer values from 4 to 30 seconds.
* `resolution` accepts `480p`, `720p`, or `1080p`.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* Reference input accepts up to 30 images, 10 videos, and 10 audio files, with at most 50 reference files in total. Reference audio requires at least one reference image or video.
* All media URLs must be publicly accessible and directly downloadable.
* Usage is billed by duration and resolution. Each reference-video duration, rounded down separately, is added to output duration for billing.
* Whether reference videos are supplied also affects the applicable billing rate.


# Seedance 2.5 Text to Video
Source: https://docs.vidgo.ai/api-manual/video-series/seedance-2-5-text-to-video

api-manual/video-series/seedance-2-5-text-to-video.json POST /api/generate/submit
Seedance 2.5 text to video model with native audio and precise motion control

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

## Models

* **seedance-2.5/text-to-video** - Seedance 2.5 text to video generation

## Input modes

* This model generates video directly from the prompt without media input.

## Notes

* `prompt`, `duration`, and `resolution` are required.
* `duration` accepts integer values from 4 to 30 seconds.
* `resolution` accepts `480p`, `720p`, or `1080p`.
* `aspect_ratio` accepts `auto`, `21:9`, `16:9`, `4:3`, `1:1`, `3:4`, or `9:16`.
* All media URLs must be publicly accessible and directly downloadable.
* Usage is billed by output duration and selected resolution.


# Vidu Q4 Image to Video
Source: https://docs.vidgo.ai/api-manual/video-series/vidu-q4-image-to-video

api-manual/video-series/vidu-q4-image-to-video.json POST /api/generate/submit
The Vidu Q4 Image to Video API animates one start image into a 3–16 second video at 540p, 720p, 1080p, 2K or 4K, keeping the image's aspect ratio.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Vidu Q4 Image to Video animates one still image into a 3–16 second video, from fast 540p drafts up to 4K. The image becomes the first frame and fixes the framing, so the output keeps the image's aspect ratio. Use it to bring illustrations, product stills or key visuals to life.

Vidu Q4 is ShengShu Technology's video model, released upstream as Vidu Q4 Preview in October 2026. It has no text-to-video mode: every request starts from images. Vidgo exposes two model IDs: [Image to Video](/api-manual/video-series/vidu-q4-image-to-video) and [Reference to Video](/api-manual/video-series/vidu-q4-reference-to-video).

## Models

* **vidu/q4/image-to-video** — Image-to-video with Vidu Q4 from one start frame, 3–16 seconds, 540p to 4K.

## Features

* Image-to-video from a single start frame.
* Optional prompt to direct motion and camera movement.
* Integer durations from 3 to 16 seconds.
* Five resolutions: `540p`, `720p`, `1080p`, `2k` and `4k`.
* Aspect ratio follows the input image.

## Required parameters

* `image_urls` (string\[]): Exactly one public HTTP(S) image URL. It is used as the first frame.

## Optional parameters

* `prompt` (string): What moves, how the camera moves and the mood. Up to 5,000 characters. Default empty.
* `duration` (integer): Output length in seconds, `3`–`16`. Default `5`.
* `resolution` (string): `540p`, `720p`, `1080p`, `2k` or `4k` (case-insensitive). Default `720p`.
* `enable_safety_checker` (boolean): Turn the upstream safety checker on or off.

## Resolution

| `resolution` | Use it for |
| - | - |
| `540p` | Fast, low-cost drafts |
| `720p` | Default; social and web clips |
| `1080p` | Full HD delivery |
| `2k` | High-resolution masters |
| `4k` | Hero shots and large screens |

The resolution sets the per-second billing rate.

## Output

The finished task returns one MP4 video in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **5-second 720p animation**: one illustration plus a short motion prompt at the default settings.
* **10-second 4K hero shot with callback**: a key visual animated at `4k` for large screens, with a `callback_url`.

## Notes

* Image-to-video does not generate audio on this ID and does not take `aspect_ratio`, `generate_audio`, `reference_image_urls` or `reference_audio_urls`; sending them returns 400. For multi-image references, voice references or native audio, use [Vidu Q4 Reference to Video](/api-manual/video-series/vidu-q4-reference-to-video).
* Vidu Q4 has no text-to-video mode. Every request needs an image.
* `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and `duration` (per output second). Reference images, reference audio and `generate_audio` do not change the rate. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.


# Vidu Q4 Reference to Video
Source: https://docs.vidgo.ai/api-manual/video-series/vidu-q4-reference-to-video

api-manual/video-series/vidu-q4-reference-to-video.json POST /api/generate/submit
The Vidu Q4 Reference to Video API builds a 3–16 second video from 1–12 reference images and up to 3 voice clips, with optional audio and up to 4K.

> [!TIP]
>
> 1. After submission, a `task_id` will be returned. If you provided a `callback_url`, when the task status becomes `finished` or `failed`, a POST request will be sent to the `callback_url`.
> 2. Regardless of whether `callback_url` is provided, you can retrieve the result through the unified [Query Task Status](/api-manual/task-management/status) endpoint.

Vidu Q4 Reference to Video builds a 3–16 second video from up to 12 reference images (characters, products, outfits, locations) and up to 3 voice clips, with optional native audio and output up to 4K. Use it for recurring characters in short series, ads that combine a product, a model and a location, and dubbed character clips.

Vidu Q4 is ShengShu Technology's video model, released upstream as Vidu Q4 Preview in October 2026. It has no text-to-video mode: every request starts from images. Vidgo exposes two model IDs: [Image to Video](/api-manual/video-series/vidu-q4-image-to-video) and [Reference to Video](/api-manual/video-series/vidu-q4-reference-to-video).

## Models

* **vidu/q4/reference-to-video** — Reference-to-video with Vidu Q4 from 1–12 images and up to 3 voice references, 3–16 seconds, 540p to 4K.

## Features

* Reference-to-video with 1–12 reference images in one request.
* Up to 3 voice references through `reference_audio_urls`.
* Prompt tags `[@reference_image_N]` and `[@reference_audio_N]` to say which reference goes where.
* Optional native audio with `generate_audio`.
* Five aspect ratios and five resolutions from `540p` to `4k`.
* Integer durations from 3 to 16 seconds.

## Required parameters

* `prompt` (string): The scene and how each reference is used. Must not be blank; up to 5,000 characters.
* `reference_image_urls` (string\[]): 1–12 public HTTP(S) image URLs.

## Optional parameters

* `reference_audio_urls` (string\[]): Up to 3 public HTTP(S) audio URLs used as voice references. Short, clean MP3 voice clips work best.
* `aspect_ratio` (string): `16:9`, `9:16`, `4:3`, `3:4` or `1:1`. When omitted, the upstream default (16:9) applies.
* `generate_audio` (boolean): Generate native audio with the video. When omitted, no audio is generated.
* `duration` (integer): Output length in seconds, `3`–`16`. Default `5`.
* `resolution` (string): `540p`, `720p`, `1080p`, `2k` or `4k` (case-insensitive). Default `720p`.
* `enable_safety_checker` (boolean): Turn the upstream safety checker on or off.

## Resolution

| `resolution` | Use it for |
| - | - |
| `540p` | Fast, low-cost drafts |
| `720p` | Default; social and web clips |
| `1080p` | Full HD delivery |
| `2k` | High-resolution masters |
| `4k` | Hero shots and large screens |

The resolution sets the per-second billing rate.

## Output

The finished task returns one MP4 video in the `files` array of [Query Task Status](/api-manual/task-management/status) and the callback payload.

## Examples

* **Two reference images, 720p**: a model and a sneaker referenced as `[@reference_image_1]` and `[@reference_image_2]`, vertical `9:16`.
* **Character with voice reference and audio, 1080p**: a character and a location, one voice clip referenced as `[@reference_audio_1]`, quoted dialogue, `generate_audio: true` and a `callback_url`.

## Notes

* References are numbered by array order: the first URL in `reference_image_urls` is `[@reference_image_1]`, the first URL in `reference_audio_urls` is `[@reference_audio_1]`.
* `image_urls` is not accepted on this ID and returns 400. To animate a single start frame, use [Vidu Q4 Image to Video](/api-manual/video-series/vidu-q4-image-to-video). Vidu Q4 has no text-to-video mode.
* More than 12 reference images or more than 3 audio references return 400.
* `model` and `callback_url` belong at the request root, not inside `input`.
* Credits are reserved at submission and depend on `resolution` and `duration` (per output second). Reference images, reference audio and `generate_audio` do not change the rate. A failed task is refunded in full. See the [pricing page](https://vidgo.ai/apis/api-pricing) for current rates.