Gemini Omni Flash pricing: Google quotes 720p; Sume lists four rows

The number Google prints

On the Gemini API pricing page (read 2026-10-05), the Gemini Omni Flash section lists output at $9.00 per million tokens for text and $17.50 per million for video, and notes billing of 5,792 tokens per second of 720p video, about $0.10 per second. The page does not break the Omni Flash video price out by 360p, 1080p or 4K; it sends you to the Veo 3.1 section for per-resolution rates.

The Google launch post for Gemini Omni 1.1 Flash (read 2026-10-05) describes four output tiers: a 360p draft mode that it says is up to 60 percent faster and a third of the cost of standard 720p, 720p standard, and 1080p or 4K premium.

What Sume lists

Sume's catalog id is gemini-omni-flash-1.1: 3 to 10 seconds, 360p, 720p, 1080p or 4K, 16:9 or 9:16, with native synced audio always on. Sume bills the provider list times 1.25 for each output second, as a function of the resolution (Video Router docs). The rates in Sume's pricing table are $0.03, $0.10, $0.15 and $0.30 per second at list, which become the Sume rows below.

Is the draft really a third?

Compare the Sume 360p and 720p rows: $0.0375 against $0.125 is 30 percent, a little under a third. That is consistent with Google's "third of the cost" wording, but it is Sume's table, not a measurement of Google's invoice. The 1.25 factor applies to both rows, so the ratio is unchanged.

One practical difference from the Google figure: $0.10 a second is the 720p list price, while Sume's 720p row is $0.125 because of the house factor. Use the Sume row when you budget a Sume wallet.

Three checks before a batch

• Send resolution explicitly instead of relying on a default, so the price you budgeted is the price you reserve.
• Native audio cannot be turned off (the API rejects generate_audio: false), so there is no cheaper silent row.
• Read the live pricing_skus from GET /v1/videos/models; the table above is a snapshot.

How to turn a token price into a clip budget

Token pricing is awkward to budget with, so convert it once. Google states 5,792 tokens per second of 720p video. At $17.50 per million tokens, a second is 5,792 x 17.50 / 1,000,000 = $0.10136, which the page rounds to about $0.10. A ten-second 720p clip is therefore about $1.01 at Google's list, and $1.25 on Sume's row, which is the clean $0.10 per second at list times 1.25.

The difference between $1.01 and $1.25 is not an error. Sume's table uses a per-second list rate and the house factor, and the Google page is describing a token approximation. For a Sume wallet, use the Sume row.

Choosing a tier

The Google post frames 360p as a draft tier and 1080p or 4K as premium. On Sume's rows a 10-second clip is $0.375 at 360p, $1.25 at 720p, $1.875 at 1080p and $3.75 at 4K. A common pattern is to draft at 360p, pick the best take, and render the final at the tier you ship. Since no seed is accepted, the final is a new generation, so budget it as a fresh job, not an upgrade.

If you only publish vertical clips for social feeds, 720p is the practical default: it is the one number Google quotes, and the middle row of Sume's table.