Google Releases Gemini Omni 1.2 Flash: Video Length Extended to 15 Seconds, Price Cut by 30%

Google DeepMind has officially released Gemini Omni 1.2 Flash, a video generation model that extends the maximum clip length from 10 to 15 seconds and cuts pricing by 30% compared to its predecessor. The model is available through multiple channels including Gemini API, AI Studio, Flow, Google Vids, and Gemini for business, with all generated videos embedded with SynthID invisible watermarks.

Technically, the new model supports clips as short as 3 seconds and significantly improves lip-sync for multi-person dialogues and stability of continuous camera movements. It also enhances consistency with reference images and materials, and introduces multi-turn conversational editing, allowing users to quickly adjust clips without regeneration. The 30% price reduction aims to lower the cost barrier for scaled video content production.

Industry-wide, this release further reduces the cost of AI video generation, and the native 15-second duration aligns closely with typical social media short video needs, potentially driving scaled adoption in advertising, marketing, and corporate training. Multi-channel availability also makes it easier for developers and enterprises to integrate the capability, intensifying competition in the video generation space.