Gemini CLI is an open-source AI agent that brings Google's Gemini models directly into developers' terminal environments. Built on TypeScript and licensed under Apache 2.0, it enables natural language-driven command-line interactions for code understanding, file manipulation, shell command execution, and automated troubleshooting. The tool connects to Gemini 2.5 Pro (and Gemini 3 models via API key) with a 1 million token context window, allowing developers to analyze entire codebases in a single session. Key capabilities include Google Search grounding for real-time web context, Model Context Protocol (MCP) support for extensibility, GEMINI.md system prompt customization, and non-interactive invocation for script integration. In 2026, Gemini CLI reached 102,000 GitHub stars and version 0.38, adding Plan Mode (enabled by default), multi-agent subagents with sandboxing, browser agent capabilities, and context compression. The free tier provides 60 model requests per minute and 1,000 requests per day with a personal Google account — making it one of the most cost-effective AI coding agents available. Gemini CLI shares underlying technology with Gemini Code Assist, ensuring a consistent AI experience across VS Code and terminal environments.