Which BPM makes a 3 to 10 second Omni clip a whole number of bars?
Gemini Omni Flash 1.1 clips run 3 to 10 seconds. At 120 BPM a 10 s clip is 5 bars; at 128 BPM no whole-second length fits. A table of the BPMs that line up.

A 10-second clip is a whole number of 4/4 bars at 60, 72, 80, 96, 120, 144 and 160 BPM, and it is not at 100, 110, 125 or 128 BPM. If you cut a Short from several Gemini Omni Flash 1.1 clips (3 to 10 seconds each on Sume) and put one generated bed under them, choosing a tempo from that list lets every cut land on a bar line.
This post is only the arithmetic and the prompt wording. Music 1.0 and the Music Router have no tempo parameter, so the tempo goes in the prompt as a number, and the Music 1.0 page warns that it is a direction, not a guaranteed value. Check the finished file against your cut list.
The formula
In 4/4 a bar lasts 240 / BPM seconds. A clip of L seconds is a whole number of bars when L divided by that bar length is an integer. The table lists the BPMs that give at least one whole-second clip length between 3 and 10, since those are the lengths the Omni catalog entry accepts.
| BPM | Bar length | Clip lengths (s) that are whole bars | Bars |
|---|---|---|---|
| 60 | 4.0 s | 4, 8 | 4 s = 1 bar, 8 s = 2 bars |
| 72 | 3.33 s | 10 | 10 s = 3 bars |
| 80 | 3.0 s | 3, 6, 9 | 3, 6, 9 s = 1, 2, 3 bars |
| 90 | 2.67 s | 8 | 8 s = 3 bars |
| 96 | 2.5 s | 5, 10 | 5 s = 2 bars, 10 s = 4 bars |
| 120 | 2.0 s | 4, 6, 8, 10 | 2, 3, 4, 5 bars |
| 144 | 1.67 s | 5, 10 | 5 s = 3 bars, 10 s = 6 bars |
| 160 | 1.5 s | 3, 6, 9 | 2, 4, 6 bars |
| 100, 110, 125, 128 | not whole | none | use a 4-bar phrase that crosses a cut |
What this costs on Sume
Omni Flash 1.1 bills per output second, and native audio is always on (the API rejects generate_audio: false). At 720p the rate is $0.125 per second, so a 10-second clip is 10 x $0.125 = $1.25 and a 5-second clip is $0.625. The bed is a flat $0.125 per generation whatever its length. A 5-second and a 10-second clip are therefore the same musical grid at 96 BPM (2 and 4 bars) and differ only in clip price.
For three clips of 10, 10 and 5 seconds at 96 BPM, the grid is 4, 4 and 2 bars: 10 bars in 25 seconds. Footage is 25 x $0.125 = $3.125 at 720p. One bed adds $0.125, for $3.25 before assembly. The prompt line is short: 96 BPM, 4/4, D minor, a break at bar 5.
Prompt and verify
Put the tempo as a number, as the seven-axis table in the docs recommends, and name one moment so the arc has a place to turn, for example breakdown to bass and claps at 0:10. At 96 BPM, 0:10 is the start of bar 5, which is exactly where a second clip begins when the first runs 10 seconds.
Then verify. Play the file and mark where bar 5 begins. If the tempo is off (a track asked for 96 BPM that really runs at 98 BPM reaches bar 5, which is 16 beats in, after 16 x 60 / 98 = 9.8 s, so the downbeat lands 0.2 s early), cut the head of the file with a Timeline audio split at the measured offset, or re-run the generation. A single generation is $0.125, so two attempts cost $0.25, the price of 2 seconds of Omni 720p footage.
Sources
Related posts
More in Models
- Packaging label text test: 12 prompts on 3 Sume models, what it costs
Test Nano Banana 2.1, Imagen 4 Ultra and GPT Image 2.5 on 12 label-text prompts first: about $1.20, $0.90 and $1.40 on Sume. A short script runs it.
- Pin lyria-3.5 or leave sume/music-auto? Read routed_model per job
sume/music-auto resolves to Lyria 3.5 today and can move without notice. When to pin lyria-3.5 on Sume and how job.request.routed_model records the engine.
- Premium image models on Sume ranked: 7 to 27 cents a picture
ChatGPT Image 2.5 high is 7 cents on Sume, Nano Banana 2.1 and Qwen Image Max 10, Nano Banana Pro 19, ChatGPT Image 2 27. Ranked, with caveats.
- 500-image job on a preview model: Hy Image 3.5 availability, plan B
OpenRouter shows 83.39% availability over 3 days for Hy Image 3.5 Preview, one provider. At 500 images that is about 83 failures to retry. Plan B on Sume.
Written by Sume