Gemini Cli

Gemini CLI is an open-source AI agent developed by Google that provides direct access to Gemini models from the command line. It is designed for developers who work primarily in the terminal, allowing them to send prompts, query and edit codebases, debug issues, and generate applications from inputs such as PDFs, images, or sketches without leaving their shell. The tool includes built-in capabilities for Google Search grounding, file system operations, shell command execution, and web fetching, and it supports conversation checkpointing and custom context files named GEMINI.md to tailor behavior to individual projects.

The project is written in TypeScript and distributed as an npm package, with installation also available through Homebrew and npx. It supports multiple authentication methods, including OAuth with a personal Google account, Gemini API keys, and Vertex AI for enterprise deployments. Extensibility is provided through Model Context Protocol (MCP) server integration, and a companion GitHub Action enables automated pull request reviews, issue triage, and on-demand assistance within repositories. The project is released under the Apache 2.0 license and publishes preview, stable, and nightly release channels.

The tool is suited to software developers and engineering teams who want AI assistance embedded in their existing command-line workflows, as well as those who need to automate tasks through non-interactive scripts. It may also interest developers evaluating Gemini models for coding tasks or building custom integrations through MCP.