Overview
- Google rolled out Gemini 3.7 Flash on Thursday, Aug. 13, and made it broadly available in Gemini Spark, the Gemini API, Google AI Studio, Android Studio, Google Antigravity, and Gemini Enterprise.
- Google’s published benchmarks show sharp gains over 3.6 Flash on coding and workflow tests, including FrontierCode rising to 43.6% from 34.4% and DeepSWE to 65.3% from 49.0%.
- The model accepts text, images, video, audio and PDFs, supports very large contexts (about 1,048,576 input tokens and 65,536 output tokens), and includes thinking_level settings to trade off latency, cost and deeper multi‑step reasoning.
- Google set an introductory API price of $0.75 per 1M input tokens and $3.75 per 1M output tokens through Dec. 31, 2026 and warns that rates will double on Jan. 1, 2027, which could materially raise run costs for long‑running agents.
- The fast three‑week cadence of Flash updates reflects a strategy to ship cheaper, fast models for coding and agents while the touted flagship Gemini 3.5 Pro remains delayed and competitors push speed and agent tiers.