Omni Flash forecast: 200, 500 or 2,000 eight-second clips a month

A monthly forecast for gemini-omni-flash-1.1 on Sume: 200, 500 or 2,000 eight-second clips cost $60 to $6,000 depending on resolution. Four-column table.

4 min readSume
All posts

A team making 200 eight-second Omni Flash clips a month spends $200.00 at 720p on Sume; 500 clips is $500.00 and 2,000 clips is $2,000.00. At 1080p those become $300.00, $750.00 and $3,000.00. The figures below are plain multiplication: clips x 8 seconds x the Sume per-second rate, where the rate is the provider list price times 1.25.

Per-clip prices

One 8-second clip costs $0.30 at 360p, $1.00 at 720p, $1.50 at 1080p and $3.00 at 4K. Those come from $0.0375, $0.125, $0.1875 and $0.375 a second. Google's Omni page lists 360p, 720p (the default), 1080p and 4K outputs and says 1080p and 4K are upscaled; Sume prices all four as separate rows.

Three monthly volumes

Pick the row that matches your volume and the column that matches your delivery size. Most teams mix columns: cheap drafts at 360p, approved shots at a higher resolution.

Monthly cost of gemini-omni-flash-1.1 on Sume, 8 seconds per clip, rates as of 2026-10-08
Volume360p720p1080p4K
200 clips (1,600 s)$60.00$200.00$300.00$600.00
500 clips (4,000 s)$150.00$500.00$750.00$1,500.00
2,000 clips (16,000 s)$600.00$2,000.00$3,000.00$6,000.00

A mixed month

Suppose 500 clips are drafts at 360p and 100 of them are re-rendered at 1080p. The drafts cost 500 x $0.30 = $150.00 and the finals cost 100 x $1.50 = $150.00, a total of $300.00 for 600 jobs. Rendering all 600 at 1080p would be $900.00. The draft-then-final habit cuts the bill by two thirds in this example, and it is the single biggest lever in the table.

What each column suits

The 360p column suits prompt testing, storyboards and client previews, where the question is whether the idea works. The 720p column suits social posts that are watched on a phone. The 1080p column suits paid ads and web pages. The 4K column suits signage and large screens, and Google's Omni page notes that 1080p and 4K are upscaled outputs, so judge a 4K sample before paying for a full month of it.

Because the per-second rate is flat inside a resolution, a 3-second clip costs 3 / 8 of the 8-second figure and a 10-second clip costs 10 / 8. If your real clips are 6 seconds, multiply the table by 0.75; if they are 10 seconds, by 1.25.

Turning the forecast into a cap

A forecast is only useful if something enforces it. Divide the monthly number by your average clip price to get a job count, and stop submitting when your own counter reaches it. For 500 clips at 720p that is 500 jobs and $500.00; if the batch script sums usage.cost from each completed poll and halts at $500.00, a prompt bug that loops cannot spend more. Run one clip first and compare its usage.cost with the per-clip figure above before you launch the rest.

Limits that shape the plan

Each job is 3 to 10 seconds, 16:9 or 9:16, with native audio that cannot be turned off. Jobs are asynchronous, so a 2,000-clip month is 2,000 submits plus polls or webhooks. Read usage.cost from each completed poll and sum it for the real number; this table is a forecast, not an invoice.

  • Rates change when the provider list changes; the Sume docs describe the x 1.25 rule, not a fixed table.
  • GET /v1/usage lists reservations, captures, refunds and top-ups, so you can reconcile a month against the forecast.

Sources

Related posts

More in Pricing

All Pricing posts

Written by Sume