Today, we're expanding the Gemini 3 model family with the release of the Gemini 3 Flash, which offers edge intelligence built for speed at a fraction of the cost. With this release, we're making the next-generation Gemini 3 intelligence accessible to everyone across all Google products.
Last month, we launched Gemini 3 in Gemini 3 Pro mode and Gemini 3 Deep Think, and the response has been overwhelming. Since launch, we have been processing more than 1T tokens per day on our API. We've seen you use Gemini 3 to learn about complex topics, build and design interactive games and understand all kinds of multimodal content.
With Gemini 3, we have introduced cross-border functionality to all complex thinking methods, multimodal perception and perception as well as agent and vibe coding functions. Gemini 3 Flash maintains this foundation, combining Gemini 3's Pro-grade thinking with Flash-level latency, performance and cost. It not only empowers daily tasks with advanced thinking, but is also our most impressive model of agent workflow.
Starting today, Gemini 3 Flash is being broadcast to millions of people around the world:
