Can't choose a video model? Send sume/auto and let Sume pick
model: sume/auto lets Sume select the video family for you: 3 to 10 seconds, 16:9 or 9:16, default 720p and 8 seconds. What it does and does not tell you.

If you cannot pick among twelve video ids, send model: "sume/auto" to POST /v1/videos and Sume selects the family. The Auto controls default to 720p and 8 seconds and accept 3 to 10 second clips at 16:9 or 9:16. The response shows sume/auto as the model; Sume does not disclose which family served the request.
Request
This is the example from the Video Router docs.
curl -X POST https://api.sume.com/v1/videos \
-H "Authorization: Bearer $SUME_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "sume/auto",
"prompt": "A vertical UGC-style product clip on a desk, natural light",
"aspect_ratio": "9:16",
"duration": 5
}'What Auto fixes and what it does not
Routing is a pure function of the normalized request and the catalog version, so an idempotent replay gets the same price and the same route.
| Question | Answer |
|---|---|
| Which duration range? | 3 to 10 seconds |
| Which aspect ratios? | 16:9 or 9:16 |
| Default resolution and length? | 720p and 8 seconds |
| Can I learn which family ran? | No; do not infer it from the output |
| Can I pin a family later? | Yes: send a catalog id instead |
When to pin instead
Use Auto while you explore and a pinned id when price or length matters. Anything past 10 seconds, or any quote you need to reproduce per model, needs a catalog id such as wan-3.0 or seedance-2.5.
Costs and holds
Auto is priced from the normalized request, and the same request gets the same price on replay. Because Sume does not disclose the family, do not build features that depend on which model ran; treat it as one service with one contract.
Sume reserves the estimated price at submit, so that amount must be free in the balance when you send the request. If it is not, the call fails with 402 insufficient_credits before any provider work starts.
Sources
Related posts
More in Developers
- Caption files: YouTube UTF-8, LinkedIn text-only SRT, Sume burns in
YouTube takes SRT, SBV and more as UTF-8; LinkedIn wants text-only SRT. Sume burns captions in and exports no SRT, so build the file from segments.
- Cartesia sonic-3-latest and sonic-3-preview aliases: fix model strings
Cartesia lists sonic-3-latest and sonic-3-preview as deprecated aliases. What each maps to, which ids Sume's TTS Router accepts, and a safe migration.
- Celery task for a Sume video: task id as the Idempotency-Key
One Celery task submits POST /v1/videos with its own task id as the Idempotency-Key and re-queues itself every 30 seconds until the job ends. No double bills.
- Cheapest Sume image model for a 21:9 banner from a reference photo
Qwen Image at $0.025 is the cheapest row listing 21:9 with reference input. Flux 2 Pro is next at $0.0375. A short script finds it live.
Written by Sume