Gemini 3 Flash has launched as part of Google's Gemini 3 model family, designed to deliver high-speed performance and advanced intelligence at a lower cost. Building on the foundation of Gemini 3 Pro and its Deep Think mode, this new model processes over 1 trillion tokens daily, enabling complex reasoning, multimodal understanding, and agentic coding tasks. It combines the advanced reasoning capabilities of its predecessor with reduced latency and efficiency, making it suitable for everyday tasks and more demanding workflows. Gemini 3 Flash shows impressive performance on several benchmarks. It achieved a 90.4% score on GPQA Diamond and 33.7% on Humanity’s Last Exam without tools, outperforming the previous Gemini 2.5 Pro model across various metrics. The pricing structure is competitive, set at $0.50 per million input tokens and $3 per million output tokens. Development capabilities are enhanced with a score of 78% on SWE-bench Verified, indicating strong coding performance suited for iterative development and high-frequency workflows. For users, Gemini 3 Flash is now the default model in the Gemini app, replacing the older version. It offers practical features like analyzing video and audio content for quick insights and creating functional apps from voice commands. In AI Mode for Search, it provides comprehensive responses and organizes information efficiently, making it easier to tackle complex queries. The rollout includes access for developers through platforms like Google AI Studio and Vertex AI, signaling a broad availability of this enhanced model.