Gemini Omni 1.2 Flash makes 15-second videos at 30% lower cost
Google's video model now lets you edit a clip by chatting with it, and 720p video costs $0.07 per second.
In brief
Google's video model now makes 15-second clips
Gemini Omni 1.2 Flash costs 30% less than version 1.1. Video at 720p is $0.07 per second. You can also edit a clip by chatting instead of starting over.
If you make short videos for work, the biggest change in Gemini Omni 1.2 Flash is the price: Google cut it 30% from Omni 1.1, so 720p video now costs $0.07 per second. Google DeepMind has released the video model, and the announcement says it is generally available on the Gemini API and Gemini for business.
Editing by chat
Multi-turn editing means you refine one clip over several messages instead of starting again. Google lists swapping a product or background, changing the lighting or weather, and switching the aspect ratio. It works on full 15-second clips.
Picture an ad set on a busy city street. You ask for rain, then for a night scene, and the actors stay put.
Longer clips
The model now makes clips up to 15 seconds long. The previous version, Omni 1.1, stopped at 10. You can set any length from 3 to 15 seconds, in one-second steps. Google says that is enough for a product explainer or a full social ad spot in a single go.
A cheaper routine
Google suggests a low-cost routine. Start with rough drafts at 360p or 720p, correct small problems with a follow-up request rather than regenerating the whole clip, and only then upscale your favorite to 1080p or 4K.
Holding on to details
You can give the model character, product or style images, motion clips up to three seconds, and audio cues. Google says faces, clothing, text on packaging and brand colors remain steady when the camera angle changes. It also claims less visual drift and fewer unwanted jump cuts in long camera moves.
Dialogue improved too, with clearer lip-sync and separate voices for several speakers.
Where to use it
The model reaches developers through Google AI Studio, businesses through the Google Cloud console, and everyone else through Flow, Google Vids and Google Ads. In Vids, 15-second clips come at no extra cost with a Google account or Workspace subscription. The Gemini app is rolling it out to Google AI Ultra subscribers. Outputs carry an invisible SynthID watermark.
Google says work continues on fast sports action, rapid script changes and complex physics. Those are still weak spots. The comparisons in the announcement are Google's own examples, not independent tests.
Source: Google DeepMind Blog