minimax/hailuo-3 on OpenRouter vs Sume minimax-h3 id
OpenRouter names the model minimax/hailuo-3. Sume uses bare catalog ids, so the matching id is minimax-h3, native 480p/768p, with 2K and 4K as priced upscales.

OpenRouter's video guide captions an example as generated with minimax/hailuo-3 (2K, 5s, 16:9, with audio). Sume never uses org/slug ids; the nearest Sume catalog id is minimax-h3, which runs natively at 480p or 768p for 5 to 15 seconds, with 2K and 4K as priced upscales.
The OpenRouter slug is from that caption. The mapping to minimax-h3 is my reading of the catalog names, not something either doc states, so confirm it in GET /v1/video-router/models. Sume facts are from the video generation docs, read 2026-10-01.
How does Sume name models?
For video, Sume's published contract uses bare catalog ids such as seedance-2 and never carries a provider-org prefix. Sending minimax/hailuo-3 is therefore an id Sume does not list, and ids not in the catalog are not found.
What does minimax-h3 offer?
The docs list minimax-h3 as accepting 5 to 15 seconds at native 480p and 768p. 768p is first-class and not 720p. 2K and 4K upscales are priced if requested.
| Item | Value |
|---|---|
| Id | minimax-h3 |
| Duration | 5-15 s |
| Native resolutions | 480p, 768p |
| 2K / 4K | priced upscales, not native |
| Audio | native stereo |
What does the OpenRouter caption's 2K mean on Sume?
On Sume, 2K is an upscale of the native render and is priced if requested; it is not a native 2K generation. Check the catalog's pricing rows before submitting.
What should I do?
Drop the minimax/ prefix, list the catalog, and send minimax-h3 with a duration of 5 or more. If you hit a not-found error, verify the id first. Resolution details are in MiniMax H3 768p is not 720p.
Sources
Related posts
More in Models
- Muse Voice Transcribe 25+ languages and code-switching vs Sume STT
Muse Voice Transcribe supports 25+ languages and code-switching in one conversation. For mixed-language audio on Sume STT, omit language_code or set one hint.
- Muse Voice Transcribe keyword biasing vs Sume STT: no vocab field
Muse Voice Transcribe has keyword biasing for names and domain terms. Sume STT has no vocabulary field: its request takes a language hint and word timings.
- Nano Banana multiple images at once: n range vs Gemini's count
Google says Gemini won't always return the exact image count you ask for in a prompt. On Sume, set n, and read each model's n range from the catalog.
- Nano Banana video to image: poster from a video via Sume stills
Google's Gemini API takes a video as context for a thumbnail or poster. Sume's images API takes image references only, so grab a still frame first.
Written by Sume