Google Releases Gemini 3.7 Flash With 50% Price Cut for Developers
Google released Gemini 3.7 Flash, updating its workhorse artificial intelligence model just three weeks after launching version 3.6 Flash. The new system targets software engineering, multi-step agent planning, and automated legal and financial document processing.
 |
| Credit: Google |
To drive developer adoption, Google introduced a 50% price reduction through the end of the year. The company set introductory rates at $0.75 per million input tokens and $3.75 per million output tokens, before standard rates resume on January 1, 2027.
In an official statement, Google detailed the target performance focus:
"3.7 Flash delivers substantial improvements across software engineering, knowledge work, and web development workflows — with an introductory price of half the original 3.6 Flash cost per million tokens."
Rather than retraining the base system from scratch, Google applied targeted algorithmic updates to the core reasoning architecture. The system maintains a 1-million token input window and produces up to 64,000 output tokens. Developers can adjust internal thinking parameters to balance speed against processing accuracy.
Internal test scores showed performance gains across engineering evaluations.
On the FrontierCode 1.1 Main test measuring production code quality, Gemini 3.7 Flash scored 43.6%, up from 34.4% for version 3.6.
 |
| Credit: Google |
On the DeepSWE software engineering benchmark, performance rose from 49% to 65.3%.
Document processing evaluations in financial and legal fields showed similar gains. The system registered 34% on the GDP.pdf benchmark, compared to 22% on the earlier release.
On AutomationBench, which measures real-world business workflows, the score reached 30.4% from 17%. Web development tests on Arena.ai's WebDev Arena showed an Elo rating of 1,588, up from 1,538.
Google explained the shift in model execution behavior:
"It thinks more diligently, putting in more effort into multi-step planning and tool calls. A more disciplined execution means less manual oversight and fewer retries across engineering workflows."
Google deployed the update across its developer platforms, including Google Antigravity, Google AI Studio, Android Studio, and Gemini Enterprise. Microsoft-owned GitHub made Gemini 3.7 Flash available to software teams through GitHub Copilot.
In a statement on the GitHub Blog, GitHub announced the immediate availability:
"Gemini 3.7 Flash is now available in GitHub Copilot. From our early testing, the model has made improvements in web and app development and agentic coding workflows over its previous version."
Individual subscribers reach the model through Gemini Spark, Google's personal agent operating in over 160 countries. The engine coordinates multi-app actions across Google Workspace tools, including document drafting, status reports, and email management.
The rapid three-week update cycle occurs as enterprise developers await updates on Google's flagship Pro series. Google did not announce a release date for Gemini 3.5 Pro, prioritizing fast-turnaround updates for automated background execution.
