We've released a demo of Gemini 2.5 for the new developer computer. This special model is available through the Gemini API and enables AI Agents that interact directly with user interactions. Built on Gemini 2.5 Pro, it handles complex tasks such as navigating websites and filling out forms, offering fast performance and other benchmarks.
We have made important updates to the AI fitming tool for AI flow. New features give people greater creative control and enhanced storytelling capabilities. Now with Veo 3.1 creators can immerse their videos with greater precision, including using multiple images to control characters and styles, stitching two different frames into a seamless video, and creating rich, integrated audio across all aspects.
We've introduced new vibe coding features in Google AI Studio. Our fresh new vibe experience lowers the barrier to building powerful apps. Now you can simply define your vision for a multimodal application, and Gemini automatically handles the complex process of animating the necessary models and APIs. We've also shared an engaging vibe primer to help people get started.
We've introduced Gemini Enterprise, the new front door for Google AI at work. CEO Sundar Pichage launched Gemini Enterprise, designed to be the “front door” of Google AI at work. Moving beyond simple conversations, Gemini Enterprise helps companies move forward by leveraging multiple Gemini models and moving them to corporate data. This secure platform helps employees build, deploy and manage AI Agents, and early adopters such as HCA Health and Best Buy are already reporting positive results.
