Today, we’re building on the progress of our widely used Flash series by introducing Gemini 3.7 Flash, our most intelligent workhorse model yet for coding and agents.
Loading comments…
Maker
kettle_dev
Visit Website
blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-3-7-flash/
Project Info
Product Keywords
Alternatives
Gemini 3.7 Flash is Google's latest workhorse model in the Flash series, designed specifically for coding and agentic workflows. It builds directly on the progress of its predecessor, Gemini 3.6 Flash, which launched just three weeks earlier. The model delivers substantial improvements across software engineering, knowledge work, and web development—all at an introductory price of half the original 3.6 Flash cost per million tokens. It's positioned as the most intelligent workhorse model yet, combining stronger reasoning with cost efficiency for production-scale deployments.
Gemini 3.7 Flash shows strong gains in debugging and issue resolution, with notably higher first-pass code accuracy. It also excels at generating production-ready code, as demonstrated by its DeepSWE v1.1 score of 65.3% compared to 49.0% for 3.6 Flash.
The model generates more functional layouts and feature-complete apps in fewer prompts. For UI generation, it shows high design adherence and parity based on reference inputs—whether a screenshot, image, or full design system.
For fields like finance, law, and biosciences, 3.7 Flash delivers improved reasoning and accuracy. It significantly outperforms 3.6 Flash on complex document processing and completes real-world business workflows more effectively, scoring 30.4% versus 17.0% on AutomationBench.
Available through the end of the year at $0.75 per 1M input tokens and $3.75 per 1M output tokens, this pricing combined with enhanced performance enables developers to scale production-ready agents cost effectively.
Gemini 3.7 Flash delivers near-frontier intelligence at half the price of its predecessor, making advanced agentic workflows accessible at scale.
Other tools you might consider
The combination of a 43.6% FrontierCode score and a 65.3% DeepSWE result at this price point is remarkable. Early customer feedback highlights performance and precision that are significantly better than 3.6 Flash at a low cost. The model also powers Gemini Spark, Google's 24/7 personal AI agent, improving tool use for Google Workspace apps and delivering better accuracy for complex, multi-skill workflows.
You're building coding agents, automating business workflows, or generating web interfaces at scale and want a model that balances intelligence with cost. It's also a strong fit if you process complex documents in knowledge-dense fields or need a reliable backbone for multi-step agent orchestration. Given the introductory pricing through year-end, it's an attractive option for teams looking to move from prototypes to production without breaking the budget.
