Google’s Gemini CLI: A Premature Reveal of a Developer-Focused AI Command-Line Tool

Google’s announcement of Gemini CLI, a new open-source project, was abruptly removed shortly after its public release on June 25. The blog post, which detailed the free command-line tool, was discovered by Meet Patel on X (formerly Twitter). The accidental release offered a glimpse into Google’s plan to make its Gemini AI models available directly inside developer terminals. The Gemini CLI provides developers, hobbyists, and AI enthusiasts with access to the Gemini 2.5 Pro model. The tool offers 60 free model requests per minute (1,000 per day), integration with Gemini Code Assist for VS Code users, real-time web grounding through Google Search, and support for extensions, scripting, automation, and system prompts. The tool allows users to generate code, debug, and experiment with different workflows, and even handle research tasks and create short videos. The open-source nature of the tool will allow developers to audit code, create extensions, and create their own workflows. The post, authored by senior Google engineers Taylor Mullen and Ryan Salva, was taken down, with the reason for its removal unclear. The cached version of the blog post gives an insight into the functionalities of the new AI agent and what to expect in the future.