Video Generation

Video generation is an asynchronous task: create a task to get a taskId, poll for the result, and receive the video URL when done. Both text-to-video and image-to-video (with reference images) are supported.

1. Create a Task#

POST https://api.smartapi.cc/api/openapi/v1/generate/video/tasks

Authorization: Bearer sk-your-key-here
Content-Type: application/json

Request Body#

Field	Type	Required	Description
`model`	string	Yes	Model id, e.g. `wan2.5`, `veo3.1`.
`prompt`	string	Yes	Text prompt.
`duration`	integer	No	Duration in seconds (per model’s supported values, e.g. 5/10/15).
`resolution`	string	No	Resolution, e.g. `480P` / `720P` / `1080P`.
`ratio`	string	No	Aspect ratio, e.g. `16:9` / `9:16`.
`seed`	integer	No	Random seed.
`referenceImageUrls`	array	No	Reference image URLs (image-to-video).
`generateAudio`	boolean	No	Whether to generate audio (model-dependent; affects billing).

Supported durations, resolutions, ratios, and audio support can be viewed as a list in the “Image/Video Call Rules” tab on the Models page, or in the Playground form.

curl https://api.smartapi.cc/api/openapi/v1/generate/video/tasks \
  -H "Authorization: Bearer $SMARTAPI_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "wan2.5",
    "prompt": "Waves crashing on rocks, slow motion, dusk",
    "duration": 5,
    "resolution": "720P"
  }'

Response#

{ "taskId": "task_abc123", "status": "processing" }

2. Query Task Result#

Poll every 5 seconds until a terminal status. Pass model as a query parameter.

GET https://api.smartapi.cc/api/openapi/v1/generate/video/tasks/{taskId}?model=wan2.5

{
  "taskId": "task_abc123",
  "status": "succeeded",
  "videoUrl": "https://.../result.mp4"
}

Status is normalized to: submitted / processing (in progress); succeeded (or a returned videoUrl) success; failed; canceled.

3. Cancel a Task#

POST https://api.smartapi.cc/api/openapi/v1/generate/video/tasks/{taskId}/cancel

{ "taskId": "task_abc123", "cancelled": true }

If you don’t want to poll manually, the Task Records page shows the status, progress and results of all image/video tasks, with online preview and download.

Billing#

Video models are billed in credits (not tokens), and billing happens when the task is created successfully. Common formulas:

PER_SECOND: total = credits-per-second × duration.
PER_SECOND_RESOLUTION: total = credits-per-second × duration × resolution multiplier.
If generateAudio is enabled and the model configures an audio multiplier, the result is multiplied again.

See per-model prices and resolution multipliers in the Models page and the Billing doc for details.