Google has introduced Gemini CLI, a command-line interface tool designed to integrate its Gemini AI models directly with local code environments. 

Gemini CLI allows developers to interact with codebases using natural language prompts, streamlining common tasks such as debugging, documentation, and code generation. The tool also supports more advanced use cases, including video generation through Google’s Veo 3 model, drafting research content via Deep Research, and retrieving real-time data using Google Search.

With this release, Google positions itself in direct competition with other AI coding assistants operating from the command line, such as OpenAI’s Codex CLI and Anthropic’s Claude Code.

The company has open-sourced Gemini CLI under the Apache 2.0 license, making it accessible for broader developer adoption. Free-tier users will be able to submit up to 60 model requests per minute and 1,000 requests per day—limits Google says exceed typical usage patterns during testing.

Gemini CLI complements Google’s wider suite of AI-assisted coding products, including Gemini Code Assist, which is tailored for integrated development environments. By expanding into terminal-based tools, the tech giant is aiming to reach developers who favour lightweight, scriptable environments over graphical user interfaces.