Google launched Gemini 3.7 Flash on August 13, replacing Gemini 3.6 Flash just three weeks after that model's release2,3,7. The company is positioning the update as its "most intelligent workhorse model" for coding and agentic workflows, with a 50% introductory price cut designed to undercut competitors on inference cost4,12.
Through the end of 2026, Gemini 3.7 Flash is priced at $0.75 per million input tokens and $3.75 per million output tokens — half the original cost of Gemini 3.6 Flash11. Starting January 1, 2027, pricing rises to $1.50 per million input tokens and $7.50 per million output tokens. MarkTechPost described the introductory rate as roughly a third the blended cost of Claude Sonnet 5 or GPT-5.6 Terra10.
Google attributes the rapid turnaround to algorithmic improvements to the model's core reasoning foundation rather than a new pretraining run, combined with developer feedback. The knowledge cutoff remains at March 2026, and the model accepts text, images, audio, and video across a 1M-token context window with up to 64K output tokens.
**Benchmark gains.** Senior Director Tulsee Doshi cited coding as the headline improvement. FrontierCode 1.1 Main rose from 34.4% to 43.6%, and DeepSWE v1.1 jumped from 49.0% to 65.3%. The WebDev Arena Elo score climbed to 1,588 from 1,538. Google says the model can generate functional layouts and feature-complete apps in fewer prompts while closely following a screenshot, image, or design system.
Document and automation benchmarks also advanced. GDP.pdf, which measures complex document processing, went from 22.0% to 34.0%. AutomationBench, testing common business workflow execution, rose to 30.4% from 17.0%. Google connects those results to finance, law, biosciences, and business workflow use cases.
**Distribution.** Developers can access Gemini 3.7 Flash through the Gemini API in Google AI Studio and Android Studio, or through Google Antigravity for agent-first workflows. Enterprises reach it via the Gemini Enterprise Agent Platform. The model is also rolling out immediately to Gemini Spark, Google's subscription-based AI agent service available to Google AI Pro and Ultra customers in more than 160 countries.
**Flagship Pro still absent.** The launch arrives without any release date for Gemini 3.5 Pro, Google's anticipated premium model6. Reuters reported that Google offered no details on when the Pro model would ship. Google had said in July that Gemini 3.5 Pro was being tested with partners and would be coming "soon".
ANALYSIS The pricing strategy is the most tactically consequential element of the release. By halving API costs through year-end, Google gives enterprise teams a multi-month window to evaluate whether the coding and automation improvements reduce total operating costs enough to justify committing production workloads to the Flash tier — before prices double in January 2027.
The three-week release cadence from 3.6 Flash to 3.7 Flash, driven by algorithmic refinement rather than retraining, signals that Google is treating its Flash line as a rapid-iteration surface while the flagship Pro model remains delayed. The competitive pressure is visible in both the speed of iteration and the explicit cost comparisons to Anthropic and OpenAI models cited in the model card.