Google has released Gemini 3.7 Flash, a refinement of the 3.6 Flash model featuring algorithmic improvements to its core reasoning foundation rather than new pretraining. The model supports text, images, audio, and video within a 1M-token context window and offers customizable thinking configurations to balance quality against cost and latency.

  • Pricing is set at $0.75 per 1M input tokens and $3.75 per 1M output tokens until December 31, 2026, which is half the rate of Gemini 3.6 Flash.
  • On FrontierCode 1.1 Main, it scores 43.6% compared to 34.4% for 3.6 Flash, and achieves 65.3% on DeepSWE v1.1.
  • It reaches an Elo of 1588 on WebDev Arena and improves GDP.pdf comprehension from 22.0% to 34.0%.
  • Access is limited to API and enterprise channels including Gemini API, Google AI Studio, and Android Studio; there are no open weights.

The model targets software engineering, document-heavy knowledge work, and web development, offering a cost-effective option for startups and regulated enterprises running always-on agents.