General availability for Gemini Omni 1.2 Flash brings two improvements at once, which Google does not often pair in one release: longer clips and a lower price.
Length tops out at 15 seconds, five more than Omni 1.1 allowed, and builders pick any duration between three and that ceiling one second at a time.
On cost, 720p output runs seven cents per second, which Google says undercuts the previous model by 30 percent. Work is previewed low, then finished at 4K from a 720p draft.
Editing replaces regeneration
Price is the headline. The capability that changes workflows is multi-turn conversational editing, where a follow-up instruction reshapes an existing clip instead of forcing a fresh generation.
Google’s own list of gains is long. References hold better across character, product and style images, short motion clips and audio cues. First and last frames agree more closely, speech sounds more natural with clearer lip-sync and support for multiple speakers, and the camera moves with more control and fewer jump cuts.
Where it ships on day one
Studio, the Gemini API, Cloud with Provisioned Throughput, enterprise API, Flow, Vids and Ads get the model at launch. It also reaches the Gemini app for Google AI Ultra subscribers.
Vids footage costs nothing extra for a Google account or Workspace subscription, and Google published no quota.
No benchmark or leaderboard figures appear in the announcement. Comparisons to Omni 1.1 come from demo examples rather than measured tests.
A cheaper video model inside an ad platform compresses the distance between generating a clip and paying to place it, which is the part rivals should read closely. Seven cents a second with conversational editing makes video a line item.
