In a significant development for developers, Google officially launched its open-source artificial intelligence tool ‘Gemini CLI (Command Line Interface)’ on June 25. This tool allows developers to utilize Google’s latest AI model, Gemini, directly in a terminal environment. It enables various tasks to be performed through natural language conversations without requiring complex commands or code.
Users can input everyday language requests such as “Explain this code,” “Fix this part,” or “Add a new feature,” and Gemini CLI will automatically analyze and write the code. Unlike traditional code interfaces, its standout feature is allowing users to work as if they are chatting with AI without needing to know technical syntax.
Gemini CLI extends beyond a simple coding assistance tool. By integrating Google’s AI video model ‘Veo 3,’ users can create videos with simple commands. It also supports writing research reports and accessing real-time information through Google search. Additionally, with MCP server connection functionality, it supports advanced tasks connected to external databases.
This multifunctional CLI is evolving into an integrated AI agent responsible for creation, exploration, and data connection, beyond mere code production.
With this launch, Google faces direct competition in the AI coding tool market with OpenAI’s Codex CLI and Anthropic’s Claude Code. The integration with VS Code (Visual Studio Code) and the Gemini Code Assist platform is praised for enhancing user convenience and work scalability.
Google distributes Gemini CLI under the very permissive ‘Apache 2.0’ open-source license, allowing anyone to use it freely and participate in tool development through platforms like GitHub. The strategy aims to expand the tech ecosystem and encourage collaboration.
To promote the adoption of Gemini CLI, Google has set an unprecedented free usage limit. All users can call the AI model up to 60 times per minute and 1,000 times a day, which Google explains is more than double the average usage of a typical developer. Free users can also utilize the Gemini 2.5 Pro model and a context window of up to 1 million tokens, making it virtually unrestricted for use.
Ryan Salva, Google’s Senior Product Director, stated, “The next decade will see such tools dominate the working methods not only of developers but also students, freelancers, and creators in general,” adding, “Everyone should have access to the same tools.”
The release of Gemini CLI signifies not just a tool launch, but a turning point in programming methodologies where AI and humans collaborate.