GOOGLE LAUNCHES GEMINI OMNI: ANY INPUT, ANY OUTPUT, VIDEO GROUNDED IN HOW THE WORLD WORKS At the same time it handed out personal AI agents and turbocharged its fastest model, Google launched something that deserves more attention than it received. Gemini Omni is a new series of models built around a single promise: any input, any output, with real-world knowledge baked in. Feed it video. Feed it audio. Feed it text and images combined. What comes back is video grounded in how the world actually works, not just interpolated frames stitched together. The model understands context, continuity and physical reality in ways that distinguish it from simple frame-generation approaches. Gemini Omni Flash is rolling out now to the Gemini app, Google Flow and YouTube Shorts. YouTube Shorts users get it free. Everyone else on Google AI Plus, Pro or Ultra plans can access it through the Gemini app and Google Flow today. The timing is not accidental. OpenAI’s Sora has had the video generation conversation largely to itself for months. Google is not ceding that ground. Gemini Omni represents a direct challenge to every video generation player in the market, launched by a company with YouTube distribution, Search reach and a cloud infrastructure that dwarfs what any startup can deploy. The creative tools market just picked up a new dominant player with nine hundred million monthly Gemini users already waiting. Keywords: Gemini Omni, Google video AI model, AI video generation 2026, Google IO 2026 announcements