Gemini CLI is an open-source AI agent developed by Google that brings the power of Gemini models directly to a developer's terminal. It allows developers to integrate AI into their command-line workflows for tasks like coding, debugging, and task automation. Gemini CLI is designed to understand codebases and assist with complex queries using natural language. It can be used for content generation, problem-solving, research, and task management.
Key features include AI agent capabilities, built-in tools (file operations, shell commands, web searching, web fetching), and extensibility through Model Context Protocol (MCP) and custom extensions. It employs a "reason and act" loop to understand tasks, create plans, and adapt to feedback. Gemini CLI supports Gemini 3 Pro and Gemini Flash models and uses an automatic routing strategy to select the best model for each request. The tool is available in its own terminal interface, VS Code via the Gemini Code Assist extension, and Google Cloud Shell.
Gemini CLI offers a free tier with limitations of 60 requests per minute and 1,000 requests per day using a personal Google account. For higher usage limits, users can opt for a Google AI Pro or Ultra plan or authenticate with a Gemini API key. The paid tiers, such as Gemini Code Assist Standard and Enterprise editions, offer increased model requests per user per day and minute.
As an open-source project under the Apache 2.0 license, Gemini CLI encourages community contributions for bug reporting, feature suggestions, documentation improvements, and code enhancements.
