🚀 Google Gemini CLI arrives!
Google has just released Gemini CLI, an open‑source AI agent that runs right in your terminal, powered by Gemini 2.5 Pro. It supports up to 1 million tokens and lets developers use natural‑language prompts for everything from writing and debugging code to doing deep research—even generating images or videos!
Key Features
Open source & flexible: Licensed under Apache 2.0. Users can inspect the code, improve it, and add features via the Model Context Protocol (MCP)
Smart terminal tool: Combines code file manipulation, shell command execution, and natural‑language assistance. It also integrates Google Search, and tools like Imagen and Veo for multimedia tasks
Free preview & generous quota: Signing in with a personal Google account grants free preview access, including up to 60 model requests per minute and 1,000 per day—claimed as the most generous allowance in the industry
VS Code pairing: Built-in integration with Gemini Code Assist, so you can seamlessly switch between IDE and CLI workflows
Why it matters
No cost for most use cases: The free tier is enough for the majority of developers
Better competition: Gemini CLI joins rivals like OpenAI’s Codex CLI and Anthropic’s Claude Code, but stands out due to its open‑source and high‑volume free usage
Extensible platform: Developers can add custom workflows, plugin support, and even Docker sandboxing, making it fitting for both hobbyist and enterprise deployments
How to get started
npm install -g @google/gemini-cli
gemini
# then sign in when prompted
What’s next?
During the preview, you get all this for free. Later, Google may offer paid tiers for teams and high-scale users, or let you connect via Vertex AI or Google Cloud AI Studio.
GitHub repo: https://github.com/google-gemini/gemini-cli
Top comments (0)