250 product photos to 5-second 1080p clips with Omni: $234.38 on Sume
250 SKUs as 5-second image-to-video clips cost $234.38 at 1080p, $156.25 at 720p and $46.88 at 360p on Sume's Omni row. A draft plan brings it to $103.13.

Turning 250 product photos into 5-second image-to-video clips on Sume's gemini-omni-flash-1.1 costs $234.38 at 1080p, $156.25 at 720p and $46.88 at 360p. That is 250 x 5 = 1,250 seconds at $0.1875, $0.125 and $0.0375 a second. A cheaper plan is to draft all 250 at 360p ($46.88) and re-render the 60 best at 1080p (60 x 5 x $0.1875 = $56.25), for $103.13 in total.
The request
Image-to-video on this model sends the product photo as the first frame. In the Video Router, image_url is the first frame and end_image_url is an optional last frame; in /v1/videos the same thing is a frame_images entry with frame_type: first_frame. Duration is 5, resolution is your choice, and the aspect ratio is 16:9 or 9:16. The photo has to be reachable over public HTTPS.
Cost of the catalog
The table has the exact figures, computed as clips x 5 seconds x rate. At 1080p that is 5 x $0.1875 = $0.9375 a clip before the job is rounded up to a whole cent.
| Plan | Seconds | Cost on Sume |
|---|---|---|
| All 250 at 360p | 1,250 | $46.88 |
| All 250 at 720p | 1,250 | $156.25 |
| All 250 at 1080p | 1,250 | $234.38 |
| All 250 at 4K | 1,250 | $468.75 |
| 250 drafts at 360p + 60 finals at 1080p | 1,550 | $103.13 |
Practical notes
Submit in waves so you do not hold the whole reserve at once; 250 jobs at 1080p reserve $234.38 together. Give each job an Idempotency-Key built from the SKU, so a rerun of the script returns the same job for a SKU instead of paying twice. Log usage.cost by SKU and reconcile against the table.
Native audio is always on for this model, and the API rejects generate_audio: false, so every clip will have sound. For a silent product page, remove the audio track afterward or choose a model that allows silent output.
What to review
Check for label drift: the model can change small text on a package. Review every final at full size, and keep the approved first-frame photo next to each clip. Google's Omni page says all generated videos include SynthID watermarking, which is useful provenance for a product catalog.
Cent rounding at 250 jobs
A 5-second 1080p job is 5 x $0.1875 = $0.9375, which Sume bills as $0.94 after rounding up to a whole cent. Across 250 jobs that is $235.00, not $234.38. A 5-second 360p draft is $0.1875, billed as $0.19, so 250 drafts are $47.50. For a catalog run, budget with the cent-rounded per-job figure, and treat the exact-rate figures in the table as the lower bound.
Sources
Related posts
More in Use cases
- 28-minute podcast episode on Sume: narration, bed, render = $4.125
A 28-minute narrated episode costs about $4.125 on Sume: two TTS requests ($1.20), one Music 1.0 bed ($0.125) and a 28-minute Timeline render ($2.80).
- Vidu Q4 takes 3 audio references; which Sume video models take one?
Sume does not list Vidu Q4 Preview. Seedance 2.x, Wan 3.0 and MiniMax H3 accept reference audio; Gemini Omni Flash 1.1 does not. Prices for 15 seconds.
- A 3-minute Short with a ducked music bed: $0.425 on Sume
YouTube Audio Library songs cap at 90 seconds in a 3-minute Short. Generate one Music 1.0 track for $0.125 and render with ducking for $0.30.
- 30 jingle candidates for 5 brands: $3.75, then trim 5 to 15 seconds
Six Music 1.0 candidates for each of five brands is 30 x $0.125 = $3.75. Trimming five winners to 15 seconds in Timeline adds $0.50. The arithmetic.
Written by Sume