Gemini 3.7 Flash

Google's new agentic workhorse in the Gemini 3 family, built on Gemini 3.6 Flash with algorithmic improvements to its reasoning foundation. The proprietary, natively multimodal model accepts text, images, video, audio, and PDFs in a 1,048,576-token context window, produces up to 65,536 tokens, and supports low, medium, and high thinking levels plus function calling, search, code execution, and computer use.

Google reports 43.6% on FrontierCode 1.1, 65.3% on DeepSWE v1.1, 85.8% on Terminal-Bench 2.1, 47.9% on OSWorld 2.0, and 53.6% on HLE-Verified. Artificial Analysis scores the highest/default checkpoint at 39 on Intelligence Index v4.3. Google launched it at $0.75 per million input tokens and $3.75 per million output tokens through December 2026; architecture, parameter count, and training scale remain undisclosed.