Now, we are expanding our Gemini 3 model family with the release of Gemini 3 Flash. It delivers frontier intelligence built for speed at a fraction of the cost. This release makes Gemini 3’s next-generation intelligence accessible to everyone across Google services.
Last month we launched Gemini 3 with Gemini 3 Pro and Gemini 3 Deep Think mode and the response was incredible. Since our release date, we have processed over 10 million tokens per day with our API. We’ve seen Gemini 3 used to vibrate code simulations to learn about complex topics, build and design interactive games, and understand all types of multimodal content.
Gemini 3 introduced cutting-edge performance across complex reasoning, multimodal and vision understanding, agent and vibe coding tasks. Gemini 3 Flash maintains this foundation, combining Gemini 3’s pro-grade inference with Flash-level latency, efficiency, and cost. Not only does it improve inference and enable everyday tasks, it is also the most impressive model of agent workflow.
Starting today, Gemini 3 Flash is rolling out to millions of people around the world.
