Google Ships Gemini 3.7 Flash Three Weeks After the Last One, at Half the Price
Google's new model targets coding and agents. Its DeepSWE score jumped from roughly 49% to 65.3%, and the introductory price is half what 3.6 Flash cost. The three-week gap between releases is a record of its own.
Google made Gemini 3.7 Flash generally available on August 13. It is built specifically for coding, agentic workflows and knowledge-heavy tasks.
The detail that stands out is not the model itself but the cadence. Only three weeks separate it from Gemini 3.6 Flash.
What the numbers say
The largest jump is on DeepSWE v1.1, the benchmark that measures whether a model can resolve real software-engineering issues. The score climbed from roughly 49% to 65.3%.
On FrontierCode 1.1, it moved from 34.4% on the previous version to 43.6%.
Google attributes the gains mostly to multi-step planning and tool use. The model is better at recognizing a dead end and changing course, and it follows instructions with higher fidelity.
The practical translation is straightforward: fewer retries, less manual correction.
Price is the real story
Introductory pricing is $0.75 per million input tokens and $3.75 per million output tokens, half what 3.6 Flash cost.
That number has an expiry date. On January 1, 2027, it doubles.
If you are building an agent that burns millions of tokens, that few-month window is worth planning around.
What it means
Last year the LLM race was about which one is smarter. This year it is turning into which one is smarter per dollar.
The Flash tier exists precisely for that front: not the strongest model on the market, but the best quality-to-cost ratio for work that has to run thousands of times a day.
The three-week release cycle carries its own message. Google is no longer waiting for the next big version. ⚡
Frequently asked questions
How much does Gemini 3.7 Flash cost?
Introductory pricing is $0.75 per million input tokens and $3.75 per million output tokens. That rate holds through the end of 2026 and doubles on January 1, 2027.
What changed between Gemini 3.6 Flash and 3.7 Flash?
Mostly coding. DeepSWE v1.1 rose from roughly 49% to 65.3%, and FrontierCode 1.1 went from 34.4% to 43.6%. The introductory price is also half what 3.6 Flash cost.
What is the Flash tier good for?
Work that runs thousands of times a day and where cost matters: coding agents, document processing, internal tooling. For the hardest single problems, the Pro tier still leads.
