Alibaba Wan 3.0 Model Studio request to a Sume /v1/videos body
Map a Model Studio wan3.0-video call (input.media, parameters, X-DashScope-Async) to POST /v1/videos with model wan-3.0. Field by field.

A Model Studio call to wan3.0-video becomes a Sume call with model: wan-3.0: the media array turns into frame_images and input_references, the parameters object is flattened, and the async header goes away because /v1/videos is always asynchronous. Sume lists one wan-3.0 id, so the standard and wan3.0-video-prime variants do not both appear.
Which Model Studio fields have a Sume home?
Each row pairs an Alibaba field with where it goes in a Sume body (read 2026-10-09).
| Model Studio | Sume /v1/videos | Note |
|---|---|---|
| model wan3.0-video or wan3.0-video-prime | model: wan-3.0 | One Sume id |
| X-DashScope-Async: enable | not needed | Every video job is async |
| input.prompt | prompt | Keep "Image 1" style labels as plain text |
| media type first_frame, last_frame | frame_images with frame_type | Wan 3.0 lists end-frame support |
| media type reference_image, reference_video, reference_audio | input_references with image_url, video_url, audio_url | 10 images, 5 videos, 5 audio |
| media type file, link | not exposed | Send page images instead |
| parameters.resolution 720P | resolution: 720p | Lowercase; 480p, 720p, 1080p |
| parameters.ratio | aspect_ratio | Wan row has no 21:9 |
| parameters.duration 2 to 30, or -1 | duration (integer) | No smart duration |
| parameters.audio | generate_audio | On by default |
| parameters.seed | not accepted | 400 unsupported_parameter |
| parameters.prompt_extend, watermark | no field | Not exposed |
What do the other differences change?
Four differences change how you write the client.
- Model Studio uses a workspace id and one of six regional hosts. Sume has one base URL,
https://api.sume.com. - Model Studio task ids and video URLs last 24 hours. A Sume job is read at
polling_url. - Pricing is a Sume list times 1.25, reserved at submit; check
GET /v1/videos/modelsfor the livepricing_skus.
What does the finished Sume body look like?
This is the body for a 6-second 720p clip with one first frame and a 16:9 frame.
{
"model": "wan-3.0",
"prompt": "Image 1 is the opening frame; she lifts the cup and smiles",
"duration": 6,
"resolution": "720p",
"aspect_ratio": "16:9",
"generate_audio": true,
"frame_images": [{
"type": "image_url",
"image_url": {"url": "https://example.com/frame.png"},
"frame_type": "first_frame"
}]
}The Model Studio fields come from Alibaba's Wan3.0 API reference; the Sume side is in the video generation docs.
Sources
Related posts
More in Developers
- Wan 3.0 workspace-id region hosts vs Sume's one API base URL
Alibaba's Wan 3.0 URLs carry a workspace id and a region host, in six regions. Sume's video API is one base URL; what to change when you port the call.
- Failed AI video jobs: Google bills successes, Sume refunds the hold
Google says Veo videos are charged only when generated. Sume holds the price at submit, captures it on success and refunds it on failure or early cancel.
- Audio detach retry: what idempotency_hit true means for the $0.01
Retrying an audio detach with the same Idempotency-Key and body returns the first job (idempotency_hit true): $0.01 once. A changed body returns 409.
- Audio job failed: read /events and /status before you retry
A failed TTS, STT or music job is terminal. Read GET /v1/jobs/{id}/events and status, fix the cause, then resubmit under a new key. A new TTS job bills again.
Written by Sume