Gemini Cli

Gemini CLI is Google's open source command line agent. It runs in the terminal and connects Gemini 3 to a local working directory, so it can query and edit large codebases, generate applications from images or PDFs, run commands and automate multistep workflows without leaving the shell. It installs as a global npm package. It suits developers who prefer the terminal to an editor extension, and it is extensible by design. Extensions bundle MCP servers, context files and custom commands into a single installable package that teaches the agent how to use a given tool. They install from any GitHub URL or local folder, templates exist for common cases, and a published gallery collects extensions from industry vendors and the open source community. The project maintains documentation, a command reference and a changelog. The plans page lists authentication routes rather than prices. A free tier works with any Google Account; Google AI Pro and Google AI Ultra subscriptions raise usage limits, with Ultra adding the highest limits and access to Deep Think; the Google Developer Program covers teams and includes Gemini Code Assist plus monthly cloud and generative AI credits with a corporate Workspace account; an AI Studio API key gives pay as you go access with higher rate limits and paid only models; and Vertex AI covers enterprise deployment under Google Cloud terms with an SLA and regional data residency. No dollar amounts appear on that page. One change matters for anyone evaluating the tool: a notice on the site states that for unpaid tier and Google One users, Gemini CLI was replaced by Antigravity CLI on June 18, 2026. Subscribers and API key users continue on Gemini CLI itself.