
Google released Gemini CLI: AI agent for code
Google company presented Gemini CLI. This is an official agent for using artificial intelligence to write code. Directly from the command line. And this tool provides access to all capabilities of the Gemini 2.5 Pro model.
Now developers can perform internet searches, process any files, automate workflows, including sending pull requests, and write code – all without leaving the terminal.
Particularly interesting is the possibility of extending functionality through MCP. You can generate images or videos directly from the terminal, connect Notion, Jira or other services.
Google also showed incredible generosity regarding limits. Absolutely free are 60 requests per minute and 1,000 per day.
This significantly exceeds competitors’ offerings – neither Codex nor Claude Code provide such conditions. All code is published on GitHub under Apache 2.0 license, which allows modifying it for your needs. Link in description.
Impressive is the context window – 1 million tokens. This means the system can process even the most voluminous projects entirely.