Gemini Omni 1.2 Flash Makes a 15-Second 720p Generation $1.05 Before Revisions

Omni 1.2 Flash adds 15-second video and conversational edits. We calculate the 720p generation cost and explain what a production trial should measure.

Google released Gemini Omni 1.2 Flash on November 5, with native video durations from three to fifteen seconds and a stated 720p price of $0.07 per second. That makes one fifteen-second generation $1.05 at the announced rate, before revisions or other separately billed operations. Google describes the rate as 30% below Omni 1.1. Launch announcement

That is a useful starting point for a budget. The more important number for a production workflow is the spend required to get a clip someone approves.

Generation cost is easy; acceptance cost needs a trial

The following arithmetic uses Google’s announced 720p rate. It is an illustrative generation budget, not an invoice estimate for every feature or platform.

Generated durationOne generationFive full generations
5 seconds$0.35$1.75
10 seconds$0.70$3.50
15 seconds$1.05$5.25

Five attempts at the longest duration therefore cost five times as much as the headline example, even if only one is kept. Human review, editing and distribution sit outside this calculation. Confirm the applicable API pricing before budgeting a job.

Editing could change the economics

The release adds conversational video editing and stronger reference following. Google says users can change details across turns while preserving more of the original scene. It also describes drafting at lower resolutions and upscaling selected results. Its launch examples are demonstrations supplied by Google; S5 Labs has not reproduced them. Editing and reference examples

Our suggested evaluation is a product clip with a fixed reference image. Ask for one change at a time, then compare the untouched areas against the previous version. Check packaging text, proportions, color, background continuity and the first and last frames. Count a revision as successful only if it fixes the requested detail without creating another rejection.

Keep the original generation, each revision and a short reason for acceptance or rejection. That record makes it possible to distinguish a cheaper request from a cheaper finished asset. The same principle applies to Images 2.6’s editing controls.

Availability is specific to the product

Google announced general availability through the Gemini API and Gemini for business. Consumer access follows separate product rollouts, including Google AI Ultra for the Gemini app. Google also acknowledges remaining difficulty with fast sports action, rapid script changes and complex physics. Availability and limitations

The Cloud model reference identifies gemini-omni-1.2-flash, supports video generation from text, images and references, and lists system instructions and context caching as unsupported. Check the Cloud model reference for the deployment surface you intend to use.

For an existing integration, review our Veo and Gemini preview migration coverage before changing model selection. Test access, request compatibility and billing together. A successful demonstration in a consumer app does not establish that your API account has the same capabilities or allowance.

Continue reading.

Insight10 min read

Gemini API Now Stores Requests From New Paid Projects for 55 Days by Default

Since October 26, new paid-tier Gemini projects log generateContent requests for 55 days unless store=false. What the toggle covers, and what it does not.

Insight16 min read

Gemini API Preview Shutdowns: Veo 3.1 Moves to Omni, Voice Follows in November

Veo 3.1 and Omni previews may end from October 22, TTS and Live previews from November 17. Cost per clip, voice replacements and the January price rise.

Insight15 min read

Nano Banana 2.1: Half the Per-Image Price, Higher Token Rates

Google's Nano Banana 2.1 about halves the per-image price but triples input rates. Nano Banana 2 is deprecated, with no shutdown date. What to check.