AI video models in the news, Oct 7 2026: which have a Sume id
Omni Flash, Veo 3.1, Kling 4.0, Ray3.2, Seedance 2.5, Wan 3.0, MiniMax H3, LTX-2.5 and Sora: which are in the Sume catalog today, and the nearest id when not.
Release trackers list a new AI video model every few weeks, and most readers want one answer: can I call it from my code today? This is the list as of 2026-10-07. The left side is what trackers report (Magic Hour and Pinggy, both read 2026-10-07). The right side is what the Sume catalog holds, checked in the repo.
Sume's video catalog has 12 ids. Ten are general generators (Grok Imagine only from an image); the other two are Motion Transfer and Recast.
Trend model against Sume catalog
| Model in the news | What is reported | On Sume | Nearest Sume id |
|---|---|---|---|
| Gemini Omni Flash | Google's default video model, first of the Gemini Omni family | yes | gemini-omni-flash-1.1 |
| Veo 3.1 | Still listed by trackers | no | gemini-omni-flash-1.1 |
| Kling 4.0 | Reported due October 2026 | no | kling-3 |
| Luma Ray3.2 | Listed by trackers | no | seedance-2.5 or gemini-omni-flash-1.1 |
| Seedance 2.5 | Jul 31 release, ~30 s native clips, top of leaderboards | yes | seedance-2.5 |
| Wan 3.0 | Top of leaderboards | yes | wan-3.0 |
| MiniMax H3 | Jul 31, 15 s, native stereo audio, open weights Aug 3 | yes | minimax-h3, minimax-h3-max |
| LTX-2.5 | Aug 11, Fast and Pro, up to 4K, open weights and API | no | gemini-omni-flash-1.1 for 4K |
| Grok Imagine | Listed by trackers | yes, image-to-video only | grok-imagine-video-1.5 |
| Sora API | Shut down | no | pick by clip length |
Reading the right-hand column
sume/auto sends ordinary text, image and reference requests to gemini-omni-flash-1.1. That is the id to start with for anything Veo 3.1 did: up to 10 seconds, 360p to 4K, 16:9 or 9:16, native audio. Where Veo 3.1 or LTX-2.5 would have given you a longer clip, Seedance 2.5 and Wan 3.0 are the only Sume ids that reach 30 seconds.
Kling 4.0 is not in the catalog, and there is no price for it to quote. Until it appears in GET /v1/videos/models, a request for it fails, so keep the model id in configuration and not in code. Kling 3 is the id that exists: 4 to 15 seconds, 720p and 1080p, start and end frames, no reference input.
A rule for new models
A model counts as available only when its id is in the live listing. Run this and read the id column. If the model you want is not there, use the nearest id in the table and write down the date you checked.
curl -s https://api.sume.com/v1/videos/models \
-H "Authorization: Bearer $SUME_API_KEY" | jq -r '.data[].id'Sources
Related posts
More in Models
- Anime-style ad with Seedance 2.5: use a character sheet as references
Send your mascot's character sheet as input_references to seedance-2.5 on Sume and test a 6-second anime-style ad at 480p before you pay for 720p.
- ByteDance models on Sume: Seedance family and Seedream prices
Seedance 2.5, 2.0, Fast and Mini plus Seedream 5 Lite, 4.5 Edit and v4, with Sume prices, limits and the pick for each job. Read 2026-10-07.
- Can Gemini Omni Flash speak my script? It takes no audio input on Sume
On Sume, Gemini Omni Flash 1.1 takes image and video references but no audio reference. It makes its own sound. For your script in a face, use avatar video.
- Same character in three AI video shots: Omni reference image
Keep one character across separate Gemini Omni clips by sending the same reference image with every request on Sume. Request shape, prompt tags, 3-shot cost.
Written by Sume