Google dropped an open-source AI agent for your terminal — free, Apache 2.0, and built on Gemini 2.5 Pro. Here's everything you need to know about the tool taking dev Twitter by storm.
Google just went full open-source with Gemini CLI — an AI-powered command-line tool that lives in your terminal and helps you write code, run commands, manage files, and ship faster. It's free, it's open source (Apache 2.0), and it uses Gemini 2.5 Pro under the hood.
If you've been sleeping on AI coding assistants because you hate leaving your terminal, this one's for you.
At its core, Gemini CLI is an agentic AI tool that runs in your shell. It can:
The key differentiator? It's fully open source. You can read the code, fork it, extend it, and trust what it's doing.
npm install -g @google/gemini-cli
gemini auth login
gemini
And you're off. Just run gemini in any project directory and start talking to it.
GCP infrastructure work is where this thing shines. Ask it to:
It has access to a 1 million token context window, so you can literally dump your entire monorepo and ask it questions.
Google is offering a generous free tier through the Gemini API for Gemini CLI users. For individual developers and learners, you can get real work done without paying anything.
This is clearly a strategic move to get developers locked into the Gemini ecosystem — and honestly? As a dev tool, it works.
Because it's Apache 2.0 licensed, the community is already building extensions. Expect MCP (Model Context Protocol) server integrations, custom tools, and IDE plugins to explode in the coming months.
If you work primarily in the terminal, build on GCP, want something free and open source, or prefer to see exactly what your AI tool is doing — yes, absolutely.
Install it tonight.