Google Ships Gemini 3.7 Flash for Coding and Agents
Just three weeks after 3.6, Google launches Gemini 3.7 Flash – stronger at coding and agents, at an intro price 50 percent below its predecessor's launch.
Illustration · AI-generated (AI IN LIFE)
At a glance
- Gemini 3.7 Flash launched August 13, 2026 – three weeks after Gemini 3.6 Flash.
- DeepSWE v1.1 rises from 49.0 to 65.3 percent; WebDev Arena Elo from 1538 to 1588.
- Intro pricing through end of 2026: $0.75/M input and $3.75/M output tokens – 50 percent below the predecessor's launch price.
- Available in Gemini Spark, AI Studio, Android Studio and Gemini Enterprise, among others.
- Per Axios, 3.7 Flash arrives before the larger Gemini 3.5 Pro.
Google has released Gemini 3.7 Flash, the next model in its Flash line – just three weeks after Gemini 3.6 Flash. Google attributes the rapid cadence to developer feedback and algorithmic innovations. The launch was covered by 9to5Google, Axios and heise, among others.
The focus is on software engineering, web development and knowledge work. The model is said to adapt better to roadblocks, clarify intent when needed and follow instructions with greater fidelity – key traits for agentic, multi-step workflows.
Benchmark numbers back the claim: on the DeepSWE v1.1 coding benchmark, Gemini rises from 49.0 to 65.3 percent, on FrontierCode 1.1 Main from 34.4 to 43.6 percent. Its WebDev Arena Elo climbs from 1538 to 1588; on knowledge-work benchmarks such as GDP.pdf it improves from 22.0 to 34.0 percent.
The pricing strategy stands out: through the end of 2026, an introductory rate of $0.75 per million input tokens and $3.75 per million output tokens applies – roughly 50 percent below the predecessor's launch price. Google is clearly targeting the fiercely contested market for coding and agent workloads.
Gemini 3.7 Flash is available via Gemini Spark (with an AI Pro/Ultra subscription), Google Antigravity, AI Studio, Android Studio and the Gemini Enterprise platform. Notably, per Axios, the Flash model arrives ahead of the larger Gemini 3.5 Pro.
FAQ
What's new in Gemini 3.7 Flash?
Clear gains on coding, web development and knowledge-work benchmarks, plus better agentic behavior: the model adapts to roadblocks and asks for clarification when intent is unclear.
What does it cost?
Through the end of 2026 an introductory price applies: $0.75 per million input tokens and $3.75 per million output tokens – about 50 percent below the predecessor's launch price.
Where can I use it?
In Gemini Spark (AI Pro/Ultra subscription), Google Antigravity, AI Studio, Android Studio and via the Gemini Enterprise platform.
