Gemini Omni 1.2 Flash: Longer Clips, Lower Price

Google DeepMind released Gemini Omni 1.2 Flash, its video generation model, through its developer API and Google Cloud. Clips can now run up to 15 seconds, up from 10 before, and the price dropped by 30% to 7 cents per second of 720p video. Users can also edit a clip step by step within a single conversation with the model, telling it what to change in the frame instead of rewriting the whole prompt from scratch.

Fifteen seconds is still too short for a finished story. But step-by-step editing within one conversation is a real time-saver. You no longer have to rebuild the whole prompt over a single detail. The 30% price cut matters here, because video generation almost always takes several attempts before you land on a final take. Assembling and timing the final cut is still a human's job. The model just speeds up the draft stage.