Google has integrated its Gemini CLI into Zed, an open-source Rust-based code editor, to enhance developer workflows by providing direct access to Gemini's AI models within the editor.

This integration enables developers to generate, refactor, and explain code instantly, as well as chat in the terminal for general questions, making coding more interactive and efficient.

Aimed at boosting productivity and collaboration, this partnership reflects Google's broader effort to embed AI into engineering workflows, supporting features like real-time AI assistance and seamless handoff back to the editor.

Developers can supply external documentation or API specifications to Gemini for advanced problem-solving, supporting extensibility through open standards.

The Gemini CLI supports standards such as MCP, markdown instructions, themes, and custom extensions, leading to rapid community-driven growth with over 1,000 pull requests merged in just eight weeks.

A 'Follow Agent' feature allows users to see AI activities in real-time as it reads and edits files, with options to review proposed changes in a pull request-like interface.

The collaboration emphasizes real-time AI assistance, including live tracking of AI actions, review interfaces, and the ability to provide external context like documentation or API specs.

The Gemini CLI and its features are now available to all users, demonstrating Google's commitment to supporting developer choice and ecosystem diversity.

Gemini CLI remains open-source on GitHub, encouraging community involvement, contributions, and shaping its future development.

Designed to be infinitely extensible, Gemini CLI facilitates integration into various development environments, supporting complex instructions and multi-file tasks with real-time progress updates.

Users can enhance AI assistance by providing additional context like URLs with documentation or API specifications, improving the AI's problem-solving capabilities.

The Gemini CLI's agent mode enables it to plan and execute code changes across entire projects, with developers able to invoke assistance for boilerplate code and optimization suggestions.

Zed offers a setting to disable all AI features for users concerned about privacy or who prefer not to use AI in their workflow.