Google launched Gemini 3.7 Flash three weeks after 3.6, with a 65.3% DeepSWE score and introductory pricing at half its predecessor's launch rate.
Key Points:
- Gemini 3.7 Flash lifted its DeepSWE v1.1 score to 65.3% and reached 1588 Elo in web development.
- Introductory pricing is $0.75 per 1M input tokens and $3.75 per 1M output tokens through 2026.
- The model is rolling out to Spark and is available across Google's developer and enterprise AI products.
Gemini 3.7 Launch
Google announced Gemini 3.7 Flash on Aug. 13, only three weeks after Gemini 3.6 Flash. The company said developer feedback and algorithmic changes drove the release, with software engineering, web development and knowledge work among its main targets. The release cadence is unusually fast.
The gains are clearest in coding. DeepSWE v1.1 reached 65.3%, up from roughly 49% for 3.6 Flash, while FrontierCode 1.1 Main rose from 34.4% to 43.6%. Web development also improved, with the WebDev Arena Elo score increasing from 1538 to 1588.
Google said the model "generates more functional layouts and feature-complete apps in fewer prompts." Knowledge-work benchmarks improved too. GDP.pdf rose from 22.0% to 34.0%, while AutomationBench increased from 17.0% to 30.4%, and Google added updated safeguards covering CBRN misuse and cyber offense.
Also Read: Anthropic Pursues A $6B Decart Deal While Preparing To Go Public
AI Performance
Outside testing showed similar gains. Niko Grupen, head of applied research at Harvey, said the model improved all-pass performance by 2.6 points over 3.6 Flash on Legal Agent Bench. Kyrill Hux, co-founder of Nunu.ai, said its benchmarks put 3.7 Flash near mid-sized frontier models at about half the cost.
Lower pricing strengthens that argument because high-volume agents can accumulate costs through repeated planning steps, tool calls and retries. Google says 3.7 Flash follows instructions more closely and handles roadblocks more effectively, which could reduce manual intervention in engineering and business workflows. The model is generally available.
Gemini 3.6 Flash launched on Jul. 21 at $1.50 per 1M input tokens and $7.50 per 1M output tokens.
Three weeks later, 3.7 Flash debuted at $0.75 and $3.75, respectively, through Dec. 31, while expanding benchmark performance and reaching Spark for AI Pro and Ultra subscribers. The quick succession extends Google's accelerated Flash update cycle.
Read Next: Virtuals Protocol Trading Volume Explodes 180%, Bulls Eye $0.68
