Gemini CLI is a free, open-source AI agent that runs inside your terminal and gives you a direct path from a prompt to Google's models. It can query and edit large codebases, generate new apps from PDFs, images, or sketches using multimodal input, and debug problems when you describe them in plain language. Built-in tools cover Google Search grounding, file operations, shell commands, and web fetching, and the context window reaches 1M tokens. Signing in with a personal Google account gives you a free tier of 60 requests per minute and 1,000 requests per day, and the code is licensed under Apache 2.0.
Developers who spend their day in a shell are the core audience: backend engineers, DevOps and platform staff, open-source maintainers, and students learning a new codebase. Because everything happens in the command line, it suits people who prefer keyboard-driven workflows over a graphical editor. Teams with stricter requirements can authenticate with an API key or Vertex AI instead of a personal account, and the npm-based install fits restricted environments through Anaconda as well.
Typical jobs include asking questions about an unfamiliar repository, planning a refactor, and tracing a bug through logs and source files. It can automate operational chores such as querying pull requests or handling complex rebases, and it runs non-interactively inside scripts for workflow automation. MCP servers connect extra capabilities, including media generation with Imagen, Veo, or Lyria. You can install it with npx, npm, Homebrew, or MacPorts, and choose preview, stable, or nightly release channels depending on how much risk you accept.
