Omni 40-second ad: resend the packshot at every 10-second step
Google extends Omni clips to 40 s in 10 s steps with 10 s of prior context. On Sume, four 720p clips with the packshot as a reference cost $5.00.

Resend the product image with every 10-second step. Google says Gemini Omni 1.1 Flash extends a clip to at most 40 seconds, in 10-second increments, and that the model can analyze up to 10 seconds of prior context. A product that appears at second 3 is therefore 30 seconds of footage away from the model's view by second 33, so give each step the packshot again.
Sume does not list an extend mode. You get the same shape with four separate 10-second gemini-omni-flash-1.1 clips, each with the packshot in reference_image_urls, joined on a Timeline. Four clips at 720p cost $5.00, plus $0.10 for the render.
What Google documents
The Google announcement gives the 40-second cumulative limit, the 10-second increments and the 10-second context window. The model page lists 3 to 10 seconds per output at 360p, 720p, 1080p and 4K, at 24 FPS, and up to 10 seconds of input for editing and extension.
Neither page says how well an object holds its look across steps. Treat drift as a risk to test, not a measured rate.
The Sume equivalent
On Sume, reference_to_video takes up to 10 reference images and up to 3 reference videos of at most 3 seconds each. You refer to them in the prompt as <IMAGE_REF_0>. Native audio is always on and generate_audio: false is rejected, so each clip brings its own sound.
Because each clip is its own job, the packshot is a fixed input that the job reads each time. That is the point of the workaround: the model never has to remember the product from 30 seconds earlier.
| Step | Seconds on screen | Reference sent | Cost at $0.125/s |
|---|---|---|---|
| 1 | 0 to 10 | packshot | $1.25 |
| 2 | 10 to 20 | packshot + last frame still | $1.25 |
| 3 | 20 to 30 | packshot | $1.25 |
| 4 | 30 to 40 | packshot | $1.25 |
| Timeline render | 40 s output | none | $0.10 |
| Total | 40 | $5.10 |
What you give up
Four independent clips are not one continuous take. There is no carried-over motion between steps, so plan cut points at natural beats and use short Timeline transitions. A native extension keeps motion flowing; the Sume route trades that for a fixed product look and four prices you can retry one at a time.
A retry of one bad 10-second step costs $1.25 at 720p. The rates above are the Sume card on 2026-10-07; read pricing_skus from GET /v1/videos/models before you budget.
Sources
Related posts
More in Media tools
- Pan and scan a 1.85:1 trailer to vertical: one crop per shot
For a 1.85:1 trailer, trim each shot, then crop it at its own x offset: width 0.304054, x 0 to 0.695946. A three-shot plan with job count, cost and order.
- Pinterest Showcase video: 3 to 60 seconds, 2:3, 1000x1500 on Sume
Pinterest's spec page lists 3-60 seconds, 2:3 and 1000 x 1500 for Showcase video. Sume Timeline can render 1000x1500 and a 3-second cut costs the same $0.10.
- Post-production cost for 100 AI clips: trim, caption, upscale
Trim is $0.02, captions $0.20, upscale $0.009 per second. Trim, caption and upscale 100 fifteen-second clips for $35.50 on Sume, then assemble them for $2.50.
- What prompt makes a 15-second Black Friday ad track?
Write the track as three timed sections: build, drop, sting. The Music Router takes it for a flat $0.125; the whole 15-second ad is $2.305 on Sume.
Written by Sume