Google has introduced Gemini CLI, a new open-source AI agent designed to integrate the powerful capabilities of its Gemini models directly into the developer’s command line interface. This initiative aims to streamline workflows and enhance productivity by bringing advanced AI assistance to the familiar and efficient terminal environment.
Announced on June 25, 2025, Gemini CLI is built upon the robust Gemini 2.5 Pro model, which boasts an impressive 1 million token context window. This allows the AI to understand and process large amounts of information, making it highly effective for complex tasks. While it is particularly adept at coding, Google emphasizes its versatility for a wide array of applications, including content generation, deep research, problem-solving, and general task management (blog.google, The Verge).
Key functionalities of Gemini CLI include code understanding, file manipulation, command execution, and dynamic troubleshooting. It offers built-in tools that allow users to ground prompts with Google Search, providing real-time external context to the model. Furthermore, its capabilities can be extended through support for the Model Context Protocol (MCP) and bundled extensions. Developers can also customize prompts and instructions to tailor Gemini to their specific needs and automate tasks by invoking the CLI non-interactively within scripts (blog.google, GitHub). The tool even supports creative use cases, such as generating short videos with Google’s Veo and Imagen AI tools.
For individual developers, Gemini CLI offers a compelling free tier. By simply logging in with a personal Google account, users can obtain a free Gemini Code Assist license. This license grants access to Gemini 2.5 Pro with an industry-leading allowance of 60 model requests per minute and 1,000 requests per day at no charge during the preview phase (blog.google, The Verge). For professional developers requiring higher usage limits or the ability to use specific models, options include leveraging a Google AI Studio or Vertex AI key for usage-based billing, or subscribing to Gemini Code Assist Standard or Enterprise licenses.
Gemini CLI is fully open source under the Apache 2.0 license, promoting transparency and community collaboration (GitHub). This open approach encourages developers worldwide to contribute by reporting bugs, suggesting features, improving security practices, and submitting code enhancements. Its design emphasizes extensibility, building on emerging standards like MCP and system prompts (via GEMINI.md), allowing for deep customization for both personal and team configurations (blog.google).
The new CLI shares its core technology with Gemini Code Assist, Google’s AI coding assistant. This means the powerful agent mode, which enables multi-step, collaborative, and reasoning-based interactions, is available across both platforms. In VS Code, Code Assist’s agent can tirelessly work on behalf of the developer to write tests, fix errors, build features, or even migrate code, intelligently planning and recovering from failed implementation paths (blog.google, cloud.google.com).
Getting started with Gemini CLI is straightforward. Developers can install it using `npm install -g @google/gemini-cli` or by running `npx https://github.com/google-gemini/gemini-cli` (GitHub, LinkedIn). It’s worth noting that the initial announcement on June 25, 2025, experienced some temporary technical glitches, with the blog post and GitHub repository briefly returning 404 errors, leading to early speculation (Reddit, Hacker News). However, these issues have since been resolved, and the resources are now fully accessible.
Gemini CLI represents a significant advancement in integrating AI directly into the developer’s daily workflow, promising increased efficiency and a more intuitive coding experience. Its open-source nature and generous free tier position it as a strong contender in the rapidly evolving landscape of AI-powered development tools.
