Veo Pricing: Veo 3.1 Preview Is Gone, Omni's 15s Clip Costs $1.05

Veo pricing in the Gemini API comes with a twist as of November 8, 2026: there is no current Veo 3 price to pay there. Google's video generation docs list the Veo 3.1 preview models as shut down on October 22, 2026, and point developers to Gemini Omni 1.2 Flash. On Google's Gemini API pricing page, the Gemini Omni Flash price is about $0.07 per second at 720p, so a full 15-second clip costs about $1.05.

Google announced Gemini Omni 1.2 Flash as generally available on November 5, 2026. The 720p rate is 30% lower than Gemini Omni 1.1, and clips now run up to 15 seconds instead of 10. Many Veo price tables still circulating online describe models the Gemini API no longer serves, which is why the numbers below come from Google's own pages.

Veo pricing after the Veo 3.1 preview shutdown

Google's Gemini API docs list three model IDs as shut down on October 22, 2026: veo-3.1-generate-preview, veo-3.1-fast-generate-preview and veo-3.1-lite-generate-preview. The same page recommends migrating to gemini-omni-1.2-flash. For teams that still want Veo, it points to the generally available (production, not preview) Veo models on the Gemini Enterprise Agent Platform.

As of November 8, 2026, neither Google's Gemini API pricing page nor its video docs list a per-second Veo price. That leaves a hole for anyone looking for a first-party Veo 3 pricing number per second. Third-party price pages, such as BenchLM's, report figures for Veo 3.1 (Lite at $0.05 per second at 720p, Fast at $0.10, Standard at $0.40), but those trace back to reseller blogs, not to Google. Read them as unverified third-party numbers for preview models that are now shut down, not as a current Veo 3 cost per second.

How much does Gemini Omni Flash cost

Gemini Omni Flash costs $12.00 per 1 million video output tokens, according to Google's pricing page, and it is sold on the paid tier only. Tokens are the units the API counts and bills. Higher resolutions use more tokens per second, so the price per second climbs with resolution:

The previous model, Gemini Omni 1.1 Flash, is billed at $17.50 per 1 million video output tokens, about $0.10 per second at 720p. Google's pricing page says that from December 3, 2026, requests to Omni 1.1 Flash route automatically to Omni 1.2 Flash and are billed at 1.2 rates. Google also offers Provisioned Throughput (capacity reserved in advance) for the new model.

Text model prices have been falling too; our explainer on LLM API pricing covers why those rates fell and how to pay less.

Gemini Omni vs Veo

In the Gemini API, Gemini Omni vs Veo is no longer a choice between two live options. Google's docs list the Veo 3.1 previews as shut down and name Omni 1.2 Flash as the migration target, while GA Veo models stay available on the Gemini Enterprise Agent Platform. Google does not publish a side-by-side comparison of the two on the pages it links developers to.

What Google and press coverage do describe is how Omni Flash works for the user. According to The Decoder, conversational editing lets you change a clip's background, lighting or aspect ratio without regenerating it, and inputs can include character images, product photos, motion clips and audio cues. Google's announcement says reference motion clips can run up to 3 seconds each, and every output carries an imperceptible SynthID watermark (Google's invisible marker for AI-made content). The Decoder also notes that Google gives no figures for its claimed improvements to drift, lip-sync and jump cuts.

For a developer, the practical difference shows up in two places: the model ID in your request and the page you check for prices. Omni 1.2 Flash rates sit on the Gemini API pricing page; Google's Gemini API pages do not list a per-second rate for the GA Veo models they point to.

If your code still calls a veo-3.1 preview ID in the Gemini API, it needs a new target: either gemini-omni-1.2-flash in the same API, or GA Veo on the Enterprise Agent Platform. For how creators are already using rival tools such as Sora and Kling, see our piece on AI video generation.

How long can Gemini Omni videos be

Gemini Omni 1.2 Flash videos can be 3 to 15 seconds long, set in one-second steps, up from a 10-second maximum in Omni 1.1, according to Google's announcement. You draft at 360p or 720p and then upscale to 1080p or 4K.

Longer clips change the math per clip, not just per second. A 10-second Omni 1.1 clip at 720p costs 10 x $0.10 = $1.00. A 15-second Omni 1.2 clip at 720p costs 15 x $0.07 = $1.05. That is about 5% more money for 50% more footage (SyntaxRadar calculation from Google's listed rates).

SyntaxRadar calculation: what a draft-then-upscale workflow costs

Because 360p costs about a ninth of the 4K rate, the cheapest habit is to iterate small and render big once. Using Google's per-second rates for 15-second clips:

The draft-first route costs roughly a quarter of rendering every take in 4K. One assumption sits under this: that the final 4K clip is billed at the listed 4K-with-upsampling rate. Google's pricing page gives rates per second but does not spell out how an upscaled draft is billed, so check your first invoice before you scale up.

The same rates scale to a minute of finished footage, which is four 15-second clips. At 360p a minute costs about $1.38 (60 x $0.023). At 720p it is $4.20 (60 x $0.07), at 1080p about $6.24 (60 x $0.104) and at 4K about $12.54 (60 x $0.209). Those are output costs only for clips you keep; every discarded take adds its own cost at whatever resolution you rendered it.

A one-minute product video, then, is a budget question about how many takes you expect to throw away, more than about the final resolution. If you expect five rejects per keeper, the draft-first route above is where most of the saving sits.

Google's headline 30% cut is the per-second 720p change, $0.10 to $0.07. Measured per token, $17.50 to $12.00, the cut is about 31%. Google's page does not explain the one-point gap between the two measures.

Is Gemini Omni Flash free

Gemini Omni Flash is not free in the Gemini API: Google's pricing page lists it on the paid tier only, with no free tier. Outside the API, the picture differs. Google's announcement says Google Vids gets 15-second clips at no additional cost with a Google account or Workspace subscription, and 15-second videos in the Gemini app are rolling out to Google AI Ultra subscribers. The model is also generally available in Gemini for business.

That split matters for who pays what. A developer building video into an app pays per second through the API from the first clip. Someone making a short clip for a slide deck in Google Vids does not see a per-second bill at all, according to Google's announcement.

Google is staging its other Gemini releases differently too; our explainer on Gemini 4 Carbon covers why that model went to defenders first.

What to watch next

December 3, 2026 is the date to mark: Omni 1.1 Flash requests start routing to Omni 1.2 Flash and billing at the lower rate. Watch, too, whether Google adds per-second Veo pricing for the GA models on the Gemini Enterprise Agent Platform to the pages it links from the Gemini API, which would give developers a first-party Veo number to compare. And watch for Google publishing figures behind its claimed gains in drift, lip-sync and jump cuts, which The Decoder notes are missing.