Google DeepMind’s Gemini Omni 1.2 Flash can generate video clips up to 15 seconds long, with multi-turn editing and a 30% price cut from version 1.1. The changes make longer, cheaper video generation more practical, while giving creators a way to revise a clip without starting again.
Google DeepMind Watch analysis
What happened
Google says Gemini Omni 1.2 Flash is generally available through the Gemini API and Google Cloud. For 720p video, the announced price is $0.07 per second. Users can choose durations from three to 15 seconds, and guide generation with reference images, short motion clips and audio cues.
The update also adds conversational editing across a full 15-second clip. Google says users can change details such as a background, lighting or aspect ratio through follow-up prompts. Its suggested workflow is to draft at 360p or 720p, make targeted edits, then upscale selected clips to 1080p or 4K. Availability across Google products is rolling out; Google says 15-second video in Google Vids is available at no additional cost, while the Gemini app rollout is for Google AI Ultra subscribers.
Why it matters
Longer clips and a lower per-second price change the arithmetic for teams making video at scale. Being able to revise one element in a clip rather than regenerate the whole thing could also save time and reduce wasted generations, though Google has not supplied independent benchmarks for those savings.
Stronger reference adherence is aimed at a persistent production headache: keeping faces, products, colours and other details consistent as a scene changes. That is useful on paper for branded content and explainers, where a wobbly logo can become the star for all the wrong reasons.
Our read
This is a meaningful step for a video model because it combines longer clips, lower stated pricing and a more iterative editing workflow. The strongest promise is not simply that the clips last five seconds longer, but that users can carry a reference through generation and refine the result in place.
Google’s demonstrations and performance claims are the company’s own. If you are considering it for production, test the things your workflow cannot afford to get wrong, especially consistency across edits, dialogue and fast action, before building a campaign around the demo reel.
What to watch
- Whether multi-turn edits preserve people, products and scene details in everyday use.
- How the $0.07-per-second 720p price compares with the cost of repeated edits and upscaling.
- When the new capabilities reach each Google product and subscription tier.
- Whether performance improves on fast action, rapid script changes and complex physics, areas Google says remain under development.
Discussion spark: For production video, which matters most: lower generation costs, longer clips, or the ability to edit without starting over?
Sources and evidence
- Gemini Omni 1.2 Flash: Fast and cost-effective with longer clips (5 November 2026, 17:05 UTC)
not affiliated with, endorsed by, or operated by Google or Google DeepMind