- Published on
AI video models hit new heights: Google releases Gemini Omni 1.2 Flash with native 15-second generation and a 30% pri...
- Authors
- Name
- DeepHorus
On November 6, Google officially launched the video generation model Gemini Omni 1.2 Flash and made it fully available on the Gemini API and Gemini for business platform. Compared with version 1.1, the new model extends native video duration to 15 seconds, supports multi-turn conversational editing, and reduces prices by 30%. The price for 720p video generation is $0.07 per second, approximately RMB 0.47.
Video generation is moving from short-clip demonstrations to production scenarios such as advertising, e-commerce, and training, with cost and controllability becoming key to deployment. Gemini Omni 1.2 Flash focuses on addressing duration, consistency, and editing efficiency issues. Google says the model achieves full general availability for the first time, relying on Google Cloud's enterprise-grade infrastructure and security capabilities.
The new model supports custom durations in 1-second increments within a range of 3 to 15 seconds, and can generate a complete story, product explanation, or social advertising clip in one go. In the workflow, users can first quickly draft at 360p or 720p, then correct details through subsequent editing without regenerating the entire clip, and finally upgrade to 1080p or 4K to reduce costs.
In terms of reference adherence, users can guide generation by combining character, product, or style images, motion clips up to 3 seconds long, and audio cues. The model can maintain consistency in faces, clothing, packaging fonts, and brand colors across different camera angles, serving scenarios such as virtual try-on, product turntable displays, and brand video advertising. From the first frame to the last frame, the model can also maintain strong consistency of characters, objects, and scenes, reducing visual drift and distortion during motion and shot transitions.
Conversational editing allows users to replace products or backgrounds, modify lighting and weather, and switch aspect ratios round by round without generating from scratch. For audio, the model can generate smoother and more natural dialogue and clearer lip-sync, and supports multi-speaker conditional control. Users can also direct complex continuous motions such as panoramic FPV fly-throughs, dramatic pull-outs, and orbiting shots, reducing jump cuts.
Currently, developers can debug prototypes and view documentation and pricing through Google AI Studio, while enterprise users can experience it in the Google Cloud console. General users can direct scenes and perform multi-turn editing in Flow, and generate 15-second clips for free in Google Vids; the Gemini App is gradually rolling out to Google AI Ultra subscribers, and Google Ads supports scaled video ad delivery. Generated content comes with a SynthID watermark.
