Google’s Omni 1.2 Flash makes 15-second AI video 30% cheaper

Google DeepMind has released Gemini Omni 1.2 Flash, a video model that now generates clips up to 15 seconds, up from 10, at $0.07 per second for 720p video, which Google says is 30% below Omni 1.1.

It is also the first time Gemini Omni is generally available, through the Gemini API and Google’s business products.

Fifteen seconds, billed by the second

Google lets users set clip length in one-second steps from 3 to 15 seconds. The model accepts reference material: character, product or style images, motion clips of up to three seconds, and audio cues. Google says faces, garments, packaging typography and brand colors stay consistent across camera angles, and that there is less visual drift between the first and last frame.

Editing now works as a conversation. A user can swap a product or background, change lighting or weather, or switch the aspect ratio turn by turn rather than regenerating the clip. Google says this works across full 15-second clips. Speech also gets an upgrade: the company promises clearer lip-sync and better separation of several speakers in one scene. It says camera moves such as continuous FPV flythroughs and orbits now come with far fewer unwanted jump cuts.

The cost workflow Google describes is draft first, then refine. Users can draft at 360p or 720p, fix details with a follow-up edit, then upscale the best result to 1080p or 4K. Outputs carry an imperceptible SynthID watermark, and Provisioned Throughput is available for enterprise customers.

Access depends on the product:

• Developers: Google AI Studio and the Gemini API.
• Businesses: the Google Cloud console and the API.
• Flow: multi-turn edits and directed scenes.
• Google Vids: 15-second clips at no additional cost with a Google account or Workspace subscription.
• Gemini app: rolling out to Google AI Ultra subscribers.
• Google Ads: for scaling video campaigns.

What the price cut means in dollars

Independent pricing trackers put Omni 1.1 at about $0.10 per second at 720p. The blog eesel.ai, working from Google’s per-token rates of $17.50 per million video output tokens, calculated roughly $0.034 per second at 360p, $0.15 at 1080p and $0.30 at 4K for that version. That figure is consistent with the 30% reduction Google quotes. Google’s post gives only the 720p price, so the new 1080p and 4K rates should be checked on the pricing page.

At the announced rate, a full 15-second clip at 720p costs about $1.05 before any edits. Under the old rate it would have cost around $1.50, though 1.1 could not produce a clip that long in one pass. According to eesel.ai, Omni 1.1 extended clips in 10-second steps up to a cumulative 40 seconds, so the new limit mostly removes the need to stitch short pieces for typical social ad lengths.

The draft-then-upscale approach matters for budgets. Because a 360p draft costs roughly a third of a 720p one at 1.1 rates, according to eesel.ai, and a conversational edit replaces a full regeneration, the effective cost of an approved clip depends less on the headline rate than on how many revisions a team needs. Google does not publish figures on how many retries the new editing saves.

The comparison with Google’s own Veo line is less clear. Aggregator sites list Veo 3.1 Fast at about $0.10 per second and standard Veo 3.1 at about $0.40, but these are third-party figures and Google’s own pricing page should be treated as the source of record.

What Google has not shown yet

Every quality claim in the announcement, from consistency to lip-sync to the absence of cuts, comes from Google’s own examples. The post offers no benchmark scores, no side-by-side tests against other vendors’ video models and no failure rates. The only comparison with the previous version is a handful of curated clips.

Google itself lists open problems: fast sports action, rapid script changes and complex physics are still being worked on. For the Gemini app, availability is limited to Google AI Ultra subscribers at first, and the post gives no date or regional breakdown for wider access. It also does not say how the “no additional cost” Vids clips are capped, if at all.

For businesses already generating product or ad video, the practical change is a longer single clip, a lower per-second rate and the ability to produce audience variants by editing one clip. Whether the consistency gains hold up on real catalogs and brand assets is something only independent testing will show.