Twelve 10 s Gemini Omni clips make a 120 s Short for $15.20
A 120-second vertical Short from twelve 10 s Gemini Omni Flash 1.1 clips at 720p costs $15.00 to generate plus $0.20 for two Timeline minutes: $15.20 on Sume.

Twelve 10-second Gemini Omni Flash 1.1 clips at 720p cost $1.25 each, so $15.00 to generate. The 120-second render is two Timeline minutes, $0.20. The total for a vertical 120-second Short is $15.20. Gemini Omni Flash 1.1 accepts 3-10 seconds a clip at 360p, 720p, 1080p or 4K in 16:9 or 9:16, and it is the default video model behind sume/auto.
The arithmetic
The 720p price of a 10-second Omni clip is 125 cents (provider list times 1.25, rounded up). A 5-second clip is 63 cents and a 3-second clip is 38 cents. Timeline 1.0 charges $0.10 for each ceil output minute.
| Short length | Clips | Generation | Timeline minutes | Total |
|---|---|---|---|---|
| 60 s | 6 | $7.50 | 1 ($0.10) | $7.60 |
| 120 s | 12 | $15.00 | 2 ($0.20) | $15.20 |
| 180 s | 18 | $22.50 | 3 ($0.30) | $22.80 |
Twelve slots and the render strategy
The Timeline render.strategy default is auto. It chunks the render when there are more than 12 segments, and Sume refuses single above 12 slots with render_strategy_unsafe. A 120-second Short from twelve clips is on the edge of the single-pass limit. An 18-clip, 180-second Short uses the chunked path, and the price is the same $0.10 for each minute.
- The slot cap for a render is 200, so clip count is not the limit here.
- Each slot needs
startvalues that increase, andvideo[0].startmust be 0. - The last slot may stop at most 0.5 seconds before the end of the spine.
Clip length changes the total a little
Gemini Omni is billed for each output second, and the rounding to cents happens for each clip. Longer clips lose less to rounding. For 120 seconds of footage, the 720p table below shows how the total moves with the clip length. The saving is small, but the number of slots (and the number of jobs to poll) changes a lot.
| Clip length | Price per clip | Clips for 120 s | Generation | With 2 Timeline minutes |
|---|---|---|---|---|
| 3 s | $0.38 | 40 | $15.20 | $15.40 |
| 5 s | $0.63 | 24 | $15.12 (24 x 63 cents = 1,512 cents) | $15.32 |
| 10 s | $1.25 | 12 | $15.00 | $15.20 |
Sound
Gemini Omni Flash 1.1 always makes synced audio, and the API rejects generate_audio: false. The Timeline program has an audio spine (audio.url, audio.parts[] or audio.mode: "silence") and an optional soundtrack. It does not list a field that keeps the sound of each clip. If you want the sound of the clips, plan for a separate path, for example a detach per clip, and test the result before you pay for twelve generations.
Check against the Shorts rules
YouTube's Shorts page (read 2026-10-08): up to three minutes, square or vertical. A 120-second vertical render is well inside this. Omni offers 9:16 directly, so you do not need a crop. If you want a 1:1 Short, Omni does not offer 1:1 (16:9 or 9:16 only), so render a 1:1 Timeline canvas with fit set to cover, contain or blur.
Sources
Related posts
More in Pricing
- Two-hour council meeting transcript: 12 Sume jobs, $1.20
Transcribing a 2-hour public meeting on Sume STT takes 12 slices at 10 cents, $1.20. MAI-Transcribe-2 lists $0.20 batch and $1.08 streaming. No speaker labels.
- Upscale a photo for $0.20 or regenerate it: the cost comparison
Sume image upscale is $0.20 per image with a 1x to 4x factor; regenerating on nano-banana-2.1 is $0.10 and on higgsfield-soul $0.005. Prices side by side.
- Video model by budget: seconds you get for $5, $20 and $100
How many seconds of video $5, $20 and $100 buy on Sume at nine model and resolution settings, from Grok Imagine at 1.25 cents a second to Seedance 2.5 1080p.
- Vidu Q4 Preview 30% off until Nov 30: plan a test batch and a fallback
Vidu takes 30% off Q4 Preview image-to-video and reference-to-video through Nov 30. Budget a test batch and keep a Sume fallback for the same first frames.
Written by Sume