Google has introduced Gemini 3.7 Flash, a new AI model designed to deliver better performance while maintaining the speed and efficiency of its Flash series.

Google described it as its most intelligent workhorse model yet for coding and AI agents. The model is aimed at handling complex tasks, including software development, knowledge work and web development.

According to a Google announcement issued on August 13, Gemini 3.7 Flash brings improvements over Gemini 3.6 Flash in areas such as coding, debugging, web development, reasoning and multi-step workflows.

Google said the model can better handle complex instructions, planning and tool use, which can help developers build AI agents that perform tasks with less manual intervention.

The model is also designed to improve the developer experience. Google said Gemini 3.7 Flash can better adapt when it faces problems, understand user intent and follow instructions more accurately.

It can also put more effort into multi-step planning and tool calls, helping reduce the need for repeated attempts during development.

Gemini 3.7 Flash is available to developers through Google Antigravity, the Gemini API via Google AI Studio and Android Studio. Enterprises can access the model through the Gemini Enterprise Agent Platform and Gemini Enterprise app.

Individuals can use it through Gemini Spark, Google’s personal AI agent, for Google AI Pro and Ultra subscribers in supported countries.

Google has also updated Gemini Spark to use Gemini 3.7 Flash. The company said the upgrade can improve Spark’s ability to use tools and handle complex tasks, including working with Google Workspace apps. 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 through December 31, 2026.