News

Google Launches Gemini CLI, an Open-Source AI Tool for Developer Terminals

• By John K. Waters
• June 25, 2025

Google on Wednesday released Gemini CLI, a new open-source command-line interface that integrates its Gemini AI models directly into developer terminals. The tool is available in preview and is designed to support coding, research, and system-level tasks using natural language prompts.

Gemini CLI connects to Google's Gemini 2.5 Pro model, offering developers a streamlined interface to write and debug code, automate workflows, and access web content through Google Search, all within the command line.

The tool is being released under the Apache 2.0 license, one of the most permissive open-source licenses. Developers can inspect the code, contribute via GitHub, and integrate the CLI with custom workflows or team configurations.

"It provides lightweight access to Gemini, giving you the most direct path from your prompt to our model," said Taylor Mullen, Senior Staff Software Engineer, and Ryan J. Salva, Senior Director of Product Management, in Google's official blog post.

Developer-Focused Capabilities
Gemini CLI allows developers to perform tasks such as reading and writing files, generating and modifying code, executing shell commands, and interacting with local and remote tools using AI. The tool also supports the Model Context Protocol (MCP) for agent interoperability.

The CLI shares its backend with Gemini Code Assist, Google's AI tool for VS Code. Prompts submitted via either interface are handled using the same underlying model and agent logic.

Users can further automate the CLI through scripting, extend its behavior with plugins, and configure it with custom instructions via GEMINI.md.

High Usage Limits for Individuals
To promote adoption, Google is offering Gemini CLI free of charge to individual users with a personal Google account. This includes access to Gemini 2.5 Pro and a 1 million-token context window — the largest publicly available for a tool of this kind.

Each account can send up to 60 requests per minute and 1,000 per day, which Google claims is the highest free allowance available for AI developer tools. Enterprise options are available via Google AI Studio or Vertex AI.

Part of a Larger Developer Push
Gemini CLI is the latest in a series of AI tools targeted at developers. Google previously introduced Gemini Code Assist, and Jules, an asynchronous AI coding assistant.

The move comes as competition intensifies in the AI developer tooling space. Rivals include GitHub Copilot, Cursor, OpenAI's Codex CLI, and Anthropic's Claude Code.

While Google emphasizes Gemini CLI's versatility beyond programming, the tool's design — terminal-native, scriptable, and open-source — positions it primarily as a coding companion.

Adoption and Caution
Google said it expects developers to contribute to the tool's evolution and to extend its capabilities through the open-source community. However, as with other AI coding agents, trust and reliability remain open questions.

A 2024 Stack Overflow survey found that only 43% of developers trust the outputs of AI code generators, citing concerns about accuracy and security.

Gemini CLI is available now for preview. Developers can begin using the tool by logging in with a Google account and downloading the installer from the official repository.