Google has launched Gemini 3.7 Flash, its latest AI model built for coding and agentic workflows. The company introduced the model on August 13, 2026, describing it as its ‘most intelligent workhorse’ for coding and agents to date.
As Google states, “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.”
The launch comes only three weeks after Gemini 3.6 Flash. Google says the new model addresses developer feedback and improves performance across several demanding workflows.
Gemini 3.7 Flash Manages Complex Coding and Agent Tasks
Gemini 3.7 Flash chiefly focuses on software engineering, web development, and knowledge-concentrated workloads. It is created to handle multi-step planning and tool calls with higher accuracy.
Notably, the model advances debugging, issue resolution, and production-ready code generation. It also aims to reduce manual oversight and repeated attempts during engineering tasks.
When it comes to performance across leading benchmarks, 3.7 Flash showed higher scores than Gemini 3.6 Flash across several coding evaluations. It scored 43.6% on FrontierCode 1.1 Main, compared with 34.4% for Gemini 3.6 Flash. On DeepSWE v1.1, 3.7 Flash scored 65.3%, compared with 49.0% for its predecessor.
Source: Google
Source: Google
Google also reported stronger results for web development. Gemini 3.7 Flash achieved an Elo score of 1588 on WebDev Arena. Gemini 3.6 Flash recorded 1538 on the same evaluation.
The latest model also showed gains in document processing and enterprise workflow automation. Google reported a 30.4% score on AutomationBench. Gemini 3.6 Flash scored 17.0% on that benchmark. Here is a detailed depiction of benchmarks achieved by 3.7 Flash:
Source: Google
Lower Pricing Can Contribute to Higher Adoption
Google is positioning Gemini 3.7 Flash as a cost-efficient option to produce AI agents. The company has introduced lower pricing to encourage wider adoption. The introductory price is $0.75 per million input tokens. Output costs $3.75 per million tokens through December 31, 2026. Nevertheless, this resembles Gemini 3.6 Flash.
Gemini 3.7 Flash Reaches Google Antigravity
Gemini 3.7 Flash is available in Google’s agent-first development environment, Google Antigravity. Significantly, the model brings stronger reasoning to complex developer workflows to the platform.
Antigravity deals with use cases involving sub-agents, multimodal understanding, and long-horizon task execution. Developers can also use the model for mobile development and automated web improvements.
Accessibility of Gemini 3.7 Flash
The latest model is available across Google products from the launch date. Developers can start exploring agent-first workflows with Gemini 3.7 Flash in Google Antigravity. It is also available on the Gemini API via Google AI Studio and Android Studio. Google has released a support guide to help developers get started with the new model in the Gemini API.
For enterprises, access to 3.7 Flash is available in the Gemini Enterprise Agent Platform and Gemini Enterprise app. Individuals can start using the model via the 24/7 personal agent in Gemini App, Gemini Spark. However, individual accessibility is restricted to Google AI Pro and Ultra subscribers across 160 countries.
The release strengthens Google’s push toward AI agents that can perform longer and more complex tasks. It also puts greater emphasis on coding performance and affordable deployment.
Never miss learning about the latest AI models and tech advancements; stay updated with WisdomPlexus.
Recommended For You:
How Search Live Powered by Gemini 3.1 Flash Live Is Changing the Way We Search with Voice AI?
