Google has introduced Gemini 3.7 Flash, its latest AI model designed for coding. The company says the new model is its “most intelligent workhorse model yet for coding and agents,” and brings major improvements over Gemini 3.6 Flash. Google claims Gemini 3.7 Flash delivers improvements across software engineering, knowledge work, and web development workflows. In comparison charts shared by the company, Gemini 3.7 Flash also scores higher than Claude Sonnet 5 across several tests. Keep reading for all the details about Gemini 3.7 Flash.
Google claims Gemini 3.7 Flash performs better at debugging, fixing issues and generating production-ready code. On the FrontierCode 1.1 Main benchmark, it scored 43.6 per cent, compared with 34.4 per cent scored by Gemini 3.6 Flash. On DeepSWE v1.1, 3.7 Flash scored 65.3 per cent, up from 49 per cent achieved by the previous model.
The model also brings improvements to web development. Google says it can create more functional layouts and feature-complete apps in fewer prompts. It can also better match a given design, screenshot or reference image.
Gemini 3.7 Flash has also improved in areas such as finance, law and biosciences. On the GDP.pdf benchmark, it scored 34 per cent, compared with 22 per cent by Gemini 3.6 Flash. On AutomationBench, it scored 30.4 per cent, compared with 17 per cent scored by the older model.
Also read: OpenAI may soon show ads on ChatGPT in India, here’s who will see them
Google also claims that Gemini 3.7 Flash “better adapts to roadblocks” and follows instructions more closely. “It thinks more diligently, putting in more effort into multi-step planning and tool calls,” the company said.
In Google’s comparison charts, Gemini 3.7 Flash also beats Claude Sonnet 5 on FrontierCode 1.1, DeepSWE v1.1, Code Arena, GDP.pdf and AutomationBench.
Today we're introducing Gemini 3.7 Flash, our most intelligent workhorse model yet for coding and agents.
— Google (@Google) August 13, 2026
This model brings substantial gains across software engineering, web development, and complex knowledge work.
Now through the end of the year, Gemini 3.7 Flash is available… pic.twitter.com/RSCBDipjKn
Gemini 3.7 Flash is available at an introductory price of $0.75 per 1 million input tokens and $3.75 per 1 million output tokens. Google says this pricing will remain through the end of the year.
Developers can access the model through Google Antigravity and the Gemini API via Google AI Studio and Android Studio. Businesses can use it through Gemini Enterprise Agent Platform and the Gemini Enterprise app. For individual users, the new AI model is available via Gemini Spark