Google unveiled Gemini 3 Flash, a new AI model that combines frontier-level intelligence with exceptional speed and efficiency, just one day ago on December 17.
Described as offering "pro-grade reasoning at Flash-level speed and lower cost," Gemini 3 Flash builds on the Gemini 3 foundation while prioritizing efficiency. It is three times faster than Gemini 2.5 Pro, uses 30% fewer tokens on average for everyday tasks, and delivers state-of-the-art performance across benchmarks.
Key highlights include multimodal capabilities for understanding video, images, and audio; advanced reasoning for complex analysis, coding, and agentic workflows; and top scores on challenging evaluations. Notably, it achieves 90.4% on GPQA Diamond, 81.2% on MMMU Pro, 33.7% on Humanity’s Last Exam (without tools), and 78% on SWE-bench Verified—outperforming predecessors like Gemini 2.5 Pro in speed, quality, and coding.
Gemini 3 Flash is now the default model in the Gemini app worldwide, replacing Gemini 2.5 Flash, and powers AI Mode in Google Search. Developers can access it via the Gemini API, Google AI Studio, Vertex AI, and more. Pricing starts at $0.50 per million input tokens and $3 per million output tokens, making it highly cost-effective.
This release expands Google's Gemini 3 family, following the November launch of Gemini 3 Pro, and underscores the company's push toward scalable, high-performance AI.