Tech giant offers unlimited daily usage of powerful AI assistant directly in terminal
Google unveiled Thursday a new open-source artificial intelligence tool that brings its Gemini AI model directly into developers’ command-line interfaces, marking another step in the company’s push to integrate AI into everyday programming workflows.
The Gemini CLI, available free on GitHub under an Apache 2.0 license, allows developers to access Google’s Gemini 2.5 Pro model without leaving their terminal environment. The tool supports a wide range of tasks beyond coding, including content generation, research and task automation.
“For developers, the command line interface isn’t just a tool; it’s home,” Google said in announcing the release.
Generous Usage Limits Set Industry Standard
The most striking feature of Google’s offering is its usage allowance. Developers can make up to 60 model requests per minute and 1,000 requests daily at no cost — figures Google claims represent “the industry’s largest allowance” for free AI access.
Users need only a personal Google account to access the service, which includes Gemini 2.5 Pro’s 1 million token context window. Professional developers requiring higher usage can opt for paid Google AI Studio or Vertex AI keys, or subscribe to Gemini Code Assist’s Standard or Enterprise tiers.
Built-in Tools Expand Functionality
The CLI comes equipped with several advanced features that distinguish it from basic AI chat interfaces. Users can ground their prompts with Google Search results, extend functionality through the Model Context Protocol, and customize instructions for specific workflows.
The tool also supports non-interactive use, allowing developers to integrate AI assistance into existing scripts and automated processes.
Integration with Existing Google Services
Google has linked the CLI with its existing Gemini Code Assist service for Visual Studio Code users. The integration allows developers to switch between terminal and IDE environments while maintaining consistent AI assistance.
Code Assist’s “agent mode” can create multi-step plans, recover from failed implementations and suggest solutions developers might not have considered, according to Google.
Open Source Approach Invites Community Development
By releasing the code under an open-source license, Google is encouraging community contributions including bug reports, feature suggestions and security improvements. The company has established GitHub channels for issues and discussions.
The extensible design builds on emerging standards including the Model Context Protocol and system prompts, allowing users to customize their AI experience.
Market Context
The launch comes as major tech companies compete to provide AI tools for developers. Microsoft’s GitHub Copilot, OpenAI’s ChatGPT, and Anthropic’s Claude all offer coding assistance, but Google’s approach of providing unlimited free access through command lines represents a distinctive strategy.
The tool is currently in preview, with Google encouraging developers to install it directly from the GitHub repository at github.com/google-gemini/gemini-cli.
The Gemini CLI is available now for download. Users must have a personal Google account to access the free tier.