Google is making a bold play for the hearts and wallets of developers.
Just three weeks after introducing Gemini 3.6 Flash, Google on Thursday announced Gemini 3.7 Flash, which the company is calling its "most intelligent workhorse model yet for coding and agents."
Google says developer feedback and algorithmic improvements informed the accelerated release, which intensifies its competition with OpenAI and Anthropic in the AI coding and agentic workflow market.
While many AI models excel at answering questions, Gemini 3.7 Flash is engineered for execution. The model demonstrates substantial gains in software engineering, knowledge work, and web development workflows, according to Google. The company claims the model better adapts to roadblocks, clarifies intent when needed, and follows instructions with greater fidelity. Google says the model puts more effort into multistep planning and tool calls, potentially reducing the need for manual oversight.
Google’s reported benchmark results show substantial gains over Gemini 3.6 Flash, although the figures have not been independently verified. On the FrontierCode 1.1 Main benchmark, which tests coding capabilities, Gemini 3.7 Flash scored 43.6% compared to 34.4% for its predecessor. It also showed significant improvement on DeepSWE v1.1, a software engineering evaluation, jumping to 65.3% from 48.6%, according to Google’s blog post.
Beyond pure coding, the model is designed to handle more functional web development, scoring 1,588 on Arena.ai's WebDev Arena Elo ranking versus 1,538 for Gemini 3.6 Flash. For knowledge-dense fields like finance, law, and biosciences, it scored 34.0% on the GDP.pdf benchmark (up from 22.0%) and reached 30.4% on AutomationBench for real-world business workflows, compared to just 17.0% previously. Benchmark gains do not necessarily translate to better reliability, latency, or cost in production.
Google cuts introductory API pricing in half
Perhaps the most striking element of this launch is the pricing. Google is offering Gemini 3.7 Flash at an introductory rate of $0.75 per million input tokens and $3.75 per million output tokens through the end of 2026, exactly half the original cost of Gemini 3.6 Flash.
Google is betting that its reported performance gains and aggressive introductory pricing will drive adoption among developers and enterprises building AI agents. For businesses running token-heavy AI agents at scale, the introductory discount could reduce near-term operating costs.
Gemini Spark gets the new model
Gemini 3.7 Flash is also being integrated immediately into Gemini Spark, Google's personal AI agent available to Google AI Pro and Ultra subscribers in over 160 countries. The company says the update improves Spark's efficiency for knowledge work with better tool use for Google Workspace apps, delivering improved accuracy and output quality for complex, multi-skill workflows.
Google says the model includes updated safeguards against misuse involving chemical, biological, radiological, and nuclear threats, as well as offensive cybersecurity activity.
Availability and caveats
The rapid release from Gemini 3.6 Flash to Gemini 3.7 Flash shows how quickly Google is updating its smaller models amid competition from OpenAI and Anthropic. However, the elephant in the room remains Gemini 3.5 Pro, Google's flagship model. In July, Google stated the Pro model was being tested with partners and would arrive "soon," but the company offered no new details on timing with yesterday’s announcement.
While the discounted introductory price is attractive, it's worth noting that the rates are only guaranteed through the end of the year. Starting January 1, 2027, prices are set to increase to $1.50 per million input tokens and $7.50 per million output tokens, a significant jump that businesses will need to factor into their long-term planning.
Developers can access Gemini 3.7 Flash today through Google's Gemini API, AI Studio, Android Studio, and Google Antigravity, while enterprises can utilize it through Google's Enterprise AI platforms.
Read more: For additional context on how Google’s lightweight AI lineup has evolved, explore five key takeaways from the launch of Gemini 3.5 Flash.