Google has launched a new open-source AI agent that brings Gemini‘s coding, content generation, and research capabilities directly to developer terminals. According to Google, the Gemini CLI provides a “fundamental upgrade to your command-line experience,” making it easier for developers to write and debug code with natural language prompts.
“It provides easier access to Gemini by giving you the most direct path from your prompts to our model,” Google said in a statement. “While the Gemini CLI is great for coding, we built it to do so much more. It’s a versatile local utility that can be used for a wide range of tasks, from content creation and problem solving to in-depth research and task management.”
Gemini CLI uses Google’s Gemini 2.5 Pro reasoning model, which currently supports a context window of 1 million tokens (the amount of information that an AI model can understand). Google says that the Gemini CLI agent is integrated with Gemini Code Assist, includes built-in Model Context Protocol (MCP) and support for Google Search, and allows developers to generate images and videos using Veo and Imagen AI tools.
The Gemini CLI is available for developer preview today and is currently available with a free Gemini Code Assist license, which can be obtained through a personal Google account. It provides Gemini CLI users with a usage limit of 60 model requests per minute and 1,000 requests per day, which Google says is the “highest limit” offered in the industry.
Google hasn’t said whether there will be anything available for users who exceed these limits, or whether the AI agent will remain free when it becomes fully available. This generosity may give it an edge over other AI coding options such as Anthropic’s Claude, GitHub’s Copilot, or even Microsoft’s similar chatbot in Windows Terminal.
