Google DeepMind · Video generation model

Gemini Omni 1.2 Flash

Gemini Omni 1.2 Flash is Google DeepMind's video generation model, now generally available on the Gemini API and Google Cloud and through Gemini for business. It generates and edits clips natively up to 15 seconds, up from 10 seconds with Omni 1.1, with duration set in one-second steps from 3 to 15 seconds. Google says it improves reference adherence (character, product and style images, motion clips of up to three seconds, and audio cues), consistency from first to last frame, lip-sync and multi-speaker dialogue, and continuous camera moves with fewer jump cuts. Clips can be refined turn by turn in conversation instead of being regenerated. Price is cut 30% from Omni 1.1 to $0.07 per second for 720p video, with drafts at 360p or 720p and upscaling to 1080p or 4K. Outputs carry a SynthID watermark and Provisioned Throughput is available. It is also available in Flow, Google Vids, Google Ads and, for Google AI Ultra subscribers, the Gemini app. Google notes that fast sports action, rapid script changes and complex physics still need work.