Wan & HappyHorse family

Alibaba’s video generation models, six in total, called through the OfoxAI video endpoint POST /v1/videos. Tasks run asynchronously: creating a task returns a task ID, which you then use to query the result.

Choosing a model

All six support text-to-video and image-to-video, and can generate a synchronized audio track (generate_audio defaults to true). adaptive keeps the input image’s aspect ratio.

The table above matches the video_attributes returned by the Models API; if model capabilities change, the API response takes precedence. Billing is per second, and the actual charge is usage.video_cost in the task result. For prices, see the model page .

Examples

The examples below show only the request that creates the task. For the full flow of querying status and downloading the result after creation, see Video API · Quickstart.

curl -X POST https://api.ofox.ai/v1/videos \
  -H "Authorization: Bearer $OFOX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "alibaba/wan-3.0",
    "prompt": "A water town in southern China at dawn, a small boat gliding slowly under a stone bridge",
    "duration": 5,
    "resolution": "720p",
    "aspect_ratio": "16:9"
  }'

curl -X POST https://api.ofox.ai/v1/videos \
  -H "Authorization: Bearer $OFOX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "alibaba/wan-3.0",
    "prompt": "The person in the frame turns and walks into the distance",
    "duration": 5,
    "resolution": "720p",
    "frame_images": [
      { "type": "image_url", "image_url": { "url": "https://example.com/first.jpg" }, "frame_type": "first_frame" }
    ]
  }'

curl -X POST https://api.ofox.ai/v1/videos \
  -H "Authorization: Bearer $OFOX_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "alibaba/wan-2.7",
    "prompt": "Keep the camera movement and restyle the footage as ink-wash animation",
    "duration": 5,
    "resolution": "720p",
    "input_references": [
      { "type": "video_url", "video_url": { "url": "https://example.com/source.mp4" } }
    ]
  }'

• frame_images and input_references cannot be sent together; sending both returns 400 references_conflict.
• Pass "generate_audio": false when you do not need an audio track.
• For the full field structure, see Create video task.

Differences from other families

• Cannot be cancelled: once a task has been submitted upstream, DELETE /v1/videos/{id} returns 400 cancel_not_supported, and the task continues generating and is billed. Confirm your parameters before submitting.
• Video URLs: in our tests, completed tasks return only unsigned_urls (temporary Alibaba Cloud OSS addresses), with no mirror_urls. Temporary addresses may expire after 24 hours, so download and save them promptly.
• Shorter durations: Wan goes down to 2 seconds and HappyHorse to 3 seconds; the minimum for other families is 4 seconds.

Remove watermark

Parameters specific to Alibaba models are passed through provider.options.dashscope:

{
  "model": "alibaba/wan-3.0",
  "prompt": "...",
  "duration": 5,
  "resolution": "720p",
  "provider": {
    "options": { "dashscope": { "watermark": false } }
  }
}

provider.options is passed only to the matching provider; other providers do not receive it. For full details, see Video provider routing.

Common errors

For the full list of error codes, see Video API error codes.

Official docs

• Alibaba Cloud Model Studio · Model service documentation 
• OfoxAI · Create video task

FAQ