Create a video

Submit a request with the unified JSON structure and receive a task ID. Pass model-specific parameters through the common fields or metadata.

POST/v1/video/generations

Create asynchronous video generation tasks based on text prompts and optional images. The interface returns task_id for subsequent query of task status and results.

Different models may support different input formats and vendor extension parameters. Please select models visible to the current API Key that support video generation, and only pass parameters supported by that model

Authorization

bearerAuth

AuthorizationBearer <token>

Use Bearer your-api-key in the Authorization Header.

In: header

Request Body

application/json

Use JSON to submit models, prompts, and optional video generation parameters.

TypeScript Definitions

Use the request body type in TypeScript.

model*string

The video model ID to be used. Please select the model that is visible to the current API Key and supports video generation.

Length1 <= length

prompt*string

Describe the video content you want to generate. It is recommended to clarify the subject, actions, scenes, shots, and style.

Length1 <= length

duration?number

Video duration, measured in seconds. Optional range depends on the selected model.

fps?integer

Frame rate of the output video. Whether supported and available depends on the selected model.

height?integer

Output video height. Can be used with width; Support depends on the selected model.

width?integer

Output video width. Can be used with height; Support depends on the selected model.

size?string

Output video size, usually in WIDTHxHEIGHT format. Available sizes depend on the selected model.

image?string

Image input used in the image-to-video can be either the image URL or Base64 data; The specific input format depends on the selected model.

metadata?

Vendor- or model-specific extension parameters, such as negative_prompt, style, quality_level, front/end frames, and callback address. These fields are not supported by all video models

n?integer

The number of videos expected to be generated. Whether support is supported and the range of options depends on the chosen model.

response_format?string

Desired response format. Optional values depend on the chosen model and channel.

seed?integer

Used to control the generation of randomness seeds. Whether support is supported depends on the chosen model.

user?string

End user identifier provided by the caller.

Response Body

`application/json`

curl -X POST "https://example.com/v1/video/generations" \  -H "Content-Type: application/json" \  -d '{    "model": "MODEL_ID",    "prompt": "An astronaut in a spacesuit walking on the moon, high quality, cinematic quality",    "size": "1920x1080",    "image": "https://example.com/input.jpg",    "duration": 5,    "metadata": {      "seed": 20231234,      "negative_prompt": "Ambiguous",      "image_tail": "https://example.com/end-frame.png"    }  }'

{
  "id": "video_123",
  "object": "video",
  "model": "MODEL_ID",
  "created_at": 1640995200,
  "task_id": "abcd1234efgh",
  "status": "processing"
}

Authorization

Request Body

Response Body

201application/json

400application/json

401application/json

403application/json

500application/json

`application/json`

`application/json`

`application/json`

`application/json`

`application/json`