Gemini Omni Flash 1.1 ranks 8th, 4th and 3rd on three AA boards
Gemini Omni Flash 1.1 is eighth on AA text-to-video, fourth on image-to-video and third on editing. On Sume it is one id at $0.125 a second at 720p.
Gemini Omni Flash 1.1 is eighth on Artificial Analysis's text-to-video board (1,113 Elo), fourth on image-to-video (1,178) and third on video editing (1,139). It is the only model on Sume that covers all three jobs through one id, gemini-omni-flash-1.1, and it costs $0.125 per second at 720p, or $7.50 per minute. Clips run 3 to 10 seconds in 16:9 or 9:16, with native audio always on.
The three ranks
Board prices are the vendors' list prices per minute.
| Board | Rank | Elo | Listed per minute | Gap to the leader |
|---|---|---|---|---|
| Text-to-video (v2.0) | 8 | 1113 | $9.12 | 43 behind Wan 3.0 (1156) |
| Image-to-video (v1.0) | 4 | 1178 | $6.00 | 17 behind H3 Max (1195) |
| Video editing (v1.1) | 3 | 1139 | $9.64 | 48 behind Wan 3.0 (1187) |
Sume's rates
Sume lists four sizes for the model. The rate is the provider list x 1.25 for each output second.
| Resolution | List per second | Sume per second | Sume per minute | 10 s clip |
|---|---|---|---|---|
| 360p | $0.03 | $0.0375 | $2.25 | $0.375 |
| 720p | $0.10 | $0.125 | $7.50 | $1.25 |
| 1080p | $0.15 | $0.1875 | $11.25 | $1.875 |
| 4K | $0.30 | $0.375 | $22.50 | $3.75 |
What the three ranks suggest
Image-to-video is the model's best board result: 17 points behind the leader, and a $6.00 board price. That is the job to try first. Text-to-video is 43 points behind the leader, which on the usual Elo scale is a 56% win chance for the leader per vote. Editing sits third, ahead of MiniMax H3 (1,119) and HappyHorse-1.0 (1,088).
The limits shape the choice more than the rank. A 10 second ceiling and two aspect ratios rule it out for 1:1 or 4:5 placements and for any shot longer than 10 seconds. The edit route has no aspect_ratio or duration input at all.
- Still to clip, 10 seconds or less, 16:9 or 9:16: try Omni at 720p, $1.25 for 10 seconds.
- Edit an existing clip with a text instruction: Omni is the only edit model on Sume.
- Anything longer than 10 seconds: use Wan 3.0 or Seedance 2.5.
Related reading
The 4K price page covers the top size, and the Wan 3.0 vs Omni comparison covers an 8 second ad.
Cost of testing all three jobs
You can test all three board jobs for under $4 at 720p. A 5 second text-to-video clip costs $0.625, a 5 second image-to-video clip $0.625 and a 5 second edit of a 5 second source $0.625. Three of each is 9 x $0.625 = $5.625.
Use GET /v1/video-router/models to read the live capabilities before you send, because the aspect ratios and lengths differ by mode. Remember that audio is always on, so a clip you want silent needs its audio removed afterwards.
Omni takes 3 to 10 second clips, 16:9 or 9:16 only. It cannot make a 1:1 clip, a clip over 10 seconds or a silent clip, since audio is always on. For those jobs use another id: Wan 3.0 or Seedance 2.5 for length, and see each model's capabilities for ratios.
The model also accepts reference images and videos, up to 10 images and 3 videos of at most 3 seconds each, so a short reference-led clip is possible on the same id.
Sources
Related posts
More in Models
- Gemini Omni rate limit: Google lists none, Sume lists them
Google's rate-limit page has no Omni or Veo row. It lists per-project RPM, TPM and RPD plus spend caps. Sume publishes per-key limits and concurrency by plan.
- GLM-5.3 reasoning cannot be turned off: cap the Sume run instead
Z.ai says GLM-5.3 always reasons, with low, high and max levels. What that means for run time and for a spend cap on a Sume agent run.
- GLM 5.3 Fast or Flash: which one does Sume list?
Sume's agent model list has GLM 5.3 Flash, not GLM 5.3 Fast. What Z.ai's GLM-5.3 page says and what the API's model field accepts.
- Does Google keep Omni and Veo prompts for 55 days?
Google's Gemini API page says prompts, context and outputs are kept 55 days for abuse checks. Veo files last 2 days. How that differs from a Sume request.
Written by Sume