TLDR: Google has officially integrated its open-source Gemini Command Line Interface (CLI) AI agent into Zed, a high-performance, Rust-based code editor. This collaboration, announced on August 27, 2025, allows developers to leverage Gemini’s advanced AI capabilities directly within the editor for tasks such as code generation, refactoring, and real-time collaboration, marking a significant step in AI-driven software development.
Google has announced the integration of its open-source Gemini Command Line Interface (CLI) AI agent directly into the Zed code editor, a move set to revolutionize developer workflows. The partnership, unveiled on August 27, 2025, brings advanced artificial intelligence capabilities to Zed, a high-performance, open-source code editor written in Rust, renowned for its multiplayer functionalities and extensibility.
Since its initial announcement in June 2025, Google’s Gemini CLI has been designed as an extensible AI agent for the terminal, with previous integrations extending to platforms like GitHub and VS Code. The decision to integrate with Zed was driven by the editor’s robust technical foundation and engaged user base, as well as the personal preference of Google’s senior director of product for developer experiences, Ryan J. Salva, and his colleague Keith Ballinger, who are avid Zed users.
Ryan J. Salva emphasized the core philosophy behind Gemini CLI, stating, “We really built Gemini CLI to be infinitely extensible.” He highlighted its diverse applications, from building GitHub Actions and custom slash commands to organizing notes and conducting research. Salva noted that while many tools seek Gemini integration, Google proactively approached Zed, recognizing its unique position in the developer ecosystem.
This integration empowers developers with a suite of AI-driven features directly within their coding environment. Users can now generate and refactor code instantly, receive immediate explanations for code snippets or error messages, and engage in natural language chat with Gemini from Zed’s integrated terminal. A key advancement is Gemini’s ability to follow complex instructions, such as adding endpoints or writing comprehensive unit tests, executing these changes across multiple files. Upon completion, Zed provides a review interface, akin to a pull request, offering granular control over every AI-generated code modification.
Zed co-founder Nathan Sobo underscored the editor’s inherent extensibility, a factor that greatly facilitated this collaboration. “As we built our initial agentic experience in Zed for editing, I was like: we should make this extensible,” Sobo remarked, anticipating a future with a multitude of specialized AI agents. The open-source nature of both Gemini CLI and Zed further fosters community contributions and customizations, paving the way for a modular AI ecosystem within the editor.
Also Read:
- Global AI Innovators Convene in Singapore for Gemini Symposium 2025
- Adobe Integrates Google’s Gemini 2.5 Flash Image Model to Enhance Generative AI Capabilities in Firefly and Express
Developers can also provide external documentation or API specifications to Gemini, allowing the AI to leverage this context for more advanced problem-solving. This integration is now available to all users, reinforcing Google’s commitment to supporting developer choice and fostering innovation across various tools and platforms. While comparisons to rival AI agents like Anthropic’s Claude Code are inevitable, Gemini CLI aims to offer distinct advantages in accessibility and depth of integration, particularly for developers who favor command-line interfaces.
