Google has introduced Gemini 3.7 Flash, marking a significant advancement in its AI model series. This release comes just three weeks after the previous version, underscoring Google’s commitment to rapid innovation in artificial intelligence.
Gemini 3.7 Flash incorporates substantial improvements based on developer feedback and algorithmic innovations. Notably, it enhances performance in software engineering, web development, and knowledge-intensive tasks. In coding, the model demonstrates significant gains in debugging and issue resolution. For instance, it improves from 49.0% to 65.3% on the DeepSWE v1.1 benchmark and from 34.4% to 43.6% on FrontierCode 1.1 Main.
In web development, Gemini 3.7 Flash generates more functional layouts and feature-complete applications with fewer prompts. Its Elo score on Arena.ai’s WebDev Arena has increased from 1538 to 1588, indicating enhanced design adherence and parity based on reference inputs such as screenshots, images, or full design systems.
The model also excels in knowledge-dense fields like finance, law, and biosciences. It outperforms its predecessor on the GDP.pdf benchmark, improving from 22.0% to 34.0%, and demonstrates greater effectiveness in completing real-world business workflows, as evidenced by an increase from 17.0% to 30.4% on AutomationBench.
Developers will find that Gemini 3.7 Flash adapts better to challenges, clarifies intent when necessary, and follows instructions with greater fidelity. It exhibits more diligent multi-step planning and tool usage, reducing the need for manual oversight and retries in engineering workflows.
On the safety front, the model includes updated safeguards against misuse in areas such as Chemical, Biological, Radiological, and Nuclear (CBRN) domains and cyber offenses, aligning with Google’s approach to bioresilience and cybersecurity.
Until the end of 2026, Gemini 3.7 Flash is available at an introductory price of $0.75 per million input tokens and $3.75 per million output tokens, which is half the price of the previous model at launch.
In the Gemini app, 3.7 Flash is rolling out to Spark, requiring an AI Pro or Ultra subscription. The model’s improvements enhance the personal agent’s efficiency in knowledge work, with better tool integration for Google Workspace applications. Additionally, it is available in Google Antigravity, AI Studio, Android Studio, Gemini Enterprise Agent Platform, and the Gemini Enterprise app.
This rapid iteration in the Gemini series highlights Google’s dedication to advancing AI capabilities. The enhancements in Gemini 3.7 Flash, particularly in coding and knowledge-intensive tasks, reflect a strategic focus on addressing real-world challenges and improving developer experiences. As AI models continue to evolve, such swift progress sets a high standard for the industry, emphasizing the importance of responsiveness to user feedback and the continuous pursuit of excellence.
