Veo 3.1 Standard is $0.40 a second at 720p and 1080p: take 1080p

On Veo 3.1 Standard, 1080p costs the same as 720p: $0.40 per second for both, with 4k at $0.60. That is on Google's Gemini API pricing page, which also lists Veo 3.1 Fast at $0.10 for 720p, $0.12 for 1080p and $0.30 for 4k, and Veo 3.1 Lite at $0.05 for 720p and $0.08 for 1080p. So on Standard there is no saving from the lower resolution, and the choice is about size and render time.

One constraint changes the arithmetic: Google's Veo page says 1080p supports only 8-second clips.

What does an 8-second clip cost on each tier?

Multiply the per-second price by 8, since that is the only duration that works at 1080p. The prices include audio, which Veo 3.1 always generates.

Why does Standard have flat pricing between 720p and 1080p?

Google's page does not say why, so treat it as a price fact only. It does mean that the cheapest Standard choice is not the smallest, and that moving a Fast or Lite draft up to Standard changes the rate more than the resolution does: 1080p goes from $0.12 on Fast to $0.40 on Standard, a jump of more than three times.

A practical consequence is a draft ladder. Test the idea on Lite at 720p for $0.05 a second, confirm the framing on Fast, and spend Standard money only on the final render at 1080p. Failed generations do not cost anything extra: Google notes that an audio processing issue can block a video and says you are only charged when a video is generated successfully.

How does that compare with Sume's Omni rows?

Sume's catalog does not list Veo, so the comparison is Google's list against another Google-family model you can call there. Sume's Video Router doc gives gemini-omni-flash-1.1 list rates of $0.03, $0.10, $0.15 and $0.30 per second at 360p, 720p, 1080p and 4K, with Sume billing list times 1.25. That is an 8-second 1080p list of $1.20, before the multiplier, with a clip length you can choose between 3 and 10 seconds instead of being locked to 8.

Two caveats on that comparison. Google's release notes say Omni's 1080p and 4K outputs are upscaled, so the 1080p row is not native. And Omni and Veo are different models with different strengths; Google's deprecations page treats Omni as the replacement for the Veo 3.1 previews, not as an identical product.

What changes on October 22?

The three preview ids named on the pricing page, veo-3.1-generate-preview, veo-3.1-fast-generate-preview and veo-3.1-lite-generate-preview, are all listed on the deprecations page with an October 22, 2026 shutdown and gemini-omni-1.1-flash as the recommended replacement. If you budget on the table above, treat it as a price list for the next few weeks and rerun the arithmetic against the replacement's rate before you commit a monthly number.