Yoga class promo video: three beats in one 10-second clip
A studio promo in 9:16 with Gemini Omni Flash: three timecoded beats in 10 seconds, calm native sound in the prompt, $1.25 at 720p on Sume.

A yoga studio promo fits in one Omni clip: 10 seconds, 9:16, three beats written with timecodes, and a calm sound bed described in the prompt. On Sume it is $1.25 at 720p ($0.125 per second) and $1.875 at 1080p. Google's Omni page, read 2026-10-08, gives the timecode syntax as [0-3s] action description, and says the model makes an audio track by default which you steer with words such as calm background music.
This works because Omni generates sound with the picture. You are not sourcing a separate track, but you cannot switch the sound off either.
A prompt with three beats
Keep each beat to one action and one camera instruction. Name the sound once, at the end, so the model does not stack effects on every beat.
[0-3s] Wide shot of an empty sunlit studio, mats in rows, slow push-in.
[3-7s] Medium shot, an instructor flows from mountain pose to a forward fold, steady camera.
[7-10s] Close-up of hands resting on knees, soft focus, warm light.
Calm ambient music, quiet breathing. No dialogue. No scene cuts beyond these three beats.What the prompt should avoid
Google's tips list "No dialogue" and "No scene cuts" as phrases that make the result more specific. Say them if you do not want speech or extra cuts. Omni has no negative-prompt field and takes no temperature setting, so the wording of the prompt is the only control you have.
- Do not ask for a named teacher's face unless it is your own photo; Google says certain recognizable people are refused.
- Do not promise a class schedule on screen; model text can be wrong. Add dates and times as captions afterward.
- Do not use the instructor's photo if it also shows children in the EEA, Switzerland or the UK; Google blocks image uploads that contain minors there.
Cost of the promo, and of a few tries
Most clips need a few tries. Budget three drafts at 360p and one final.
| Step | Resolution | Per second | Clip cost |
|---|---|---|---|
| Draft x3 | 360p | $0.0375 | $1.125 |
| Final x1 | 720p | $0.125 | $1.25 |
| Final x1 | 1080p | $0.1875 | $1.875 |
| Draft x3 plus 1080p final | mixed | n/a | $3.00 |
Watermark and disclosure
Every Omni video carries SynthID, an invisible watermark Google says is designed to survive cropping, filters, frame-rate changes and lossy compression. The promo is not visibly marked. If your local rules require a disclosure for AI-made footage, add a caption yourself; the invisible mark does not do that job.
Making a set for the week
A studio usually needs a promo per class type. Keep the three-beat structure and change the middle beat: a flow for vinyasa, a held pose for restorative, a stretch for beginners. Five promos at 720p cost $6.25 on Sume.
Submit with an Idempotency-Key header so a retry after a dropped connection returns the original job instead of a second charge; the same key with a different payload returns 409. Then poll GET /v1/jobs/:id/status and fetch GET /v1/jobs/:id/result when it finishes.
Sources
Related posts
More in Use cases
- YouTube AI disclosure: a clip of someone giving advice they never gave
YouTube's disclosure page lists making someone appear to give advice they never gave. Its examples in a table, and what Sume trims and captions change.
- Shorts ads: Google says sound lifts conversions 20%+, add a bed
Google's Shorts ad page says sound lifts conversions by over 20%. Add a music bed that ducks under a voiceover with the Sume Timeline soundtrack field.
- Tutorial b-roll with Omni: six 10-second inserts for $7.50
One minute of b-roll for a how-to video is six Omni inserts of 10 seconds: $7.50 at 720p on Sume. Veo 3.1 clips run 4, 6 or 8 seconds, so the count changes.
- AI album cover generator: square art at 3000×3000
Generate square album art, then upscale: Apple recommends at least 3000×3000. On Sume, generate 2400×2400 and upscale it 1.25× to reach 3000×3000.
Written by Sume