Gemini Omni | I/O 2026 Keynote

TL;DR: Google announces Gemini Omni, a model that generates any output from any input, combining world knowledge with generative media for realistic video creation and editing via natural language. ## From Multimodal to World Model Over the past year, AI capabilities have leaped forward. Now we have agents that can plan and act on our behalf, and we’re close to achieving general artificial intelligence (AGI). Last year, we outlined our vision of extending Gemini’s multimodal capabilities into a world model — an AI that can understand and simulate the world. This is a key aspect of achieving AGI and is critical for everything from building AI assistants to training robots. ## Introducing Gemini Omni Now we’re taking the next big step. I’m excited to announce **Gemini Omni** — our new model that generates **any output from any input**. It combines Gemini’s intelligence with our best generative media models, enabling new levels of world understanding, multimodality, and editing. Models like Veo, Nano Banana, and Genie can produce incredibly realistic videos, images, and interactive simulations. While not perfect, they’ve already shown impressive concepts in intuitive physics. With Omni, we’ve pushed even further — this is a step change in simulating quantities like kinetic energy and gravity, which previous systems found difficult. ## World Knowledge-Driven Video Generation Gemini’s world knowledge and reasoning really shine in Omni. It translates complex concepts into highly accurate videos. For example, you can give it a simple prompt like “Make a claymation explainer video about protein folding” and get this: > (Narrator in claymation video): A protein starts as a chain of amino acids. They fold into patterns like alpha helices and flat areas called beta sheets, forming a perfect three-dimensional shape. But the initial generation is just the start. The creative process is rarely one-shot; it’s usually iterative. Just as Nano Banana redefined image editing, Omni gives you a more natural way to **edit video through conversational language**. ## Conversational Video Editing & Creative Morphing What’s really cool is that you can give it your own video — say, a selfie — and alter reality in a fun way. Adjust details and style, add elements, and the whole scene morphs to reflect your new idea. A simple circle turns into a black hole, or a night walk comes to life. Anything can become the canvas for creating a completely new reality. ## From Video to Any Modality Let’s look at what Omni can do. We started with video, but over time Omni will be able to **generate any output from any input**. That’s always been our goal for Gemini, and why we built it multimodal from the start. It was a hard road, but the results now prove the value of that foundation. ## First Model: Gemini Omni Flash Today we’re launching the first model in the Omni family: **Gemini Omni Flash**. It’s already available across our products — you’ll hear more about that shortly. We’re excited about the progress we’ve made, and soon we’ll share more about **Omni Pro**. We can’t wait to see what you create. --- Source: Gemini Omni | I/O 2026 Keynote - YouTube (https://www.youtube.com/watch?v=QhdEJFFaig0)