Some developers prefer to live in the command line interface (CLI), avoiding graphic features and management of Flashy IDE files. Google’s latest AI tool concerns these terminal enthusiasts. His name is Gemini Cli, and he shares a lot with Gemini Code Assist, but he works in your terminal environment instead of integrating into an FDI. And maybe the best of all, it’s free and open source.
Gemini Cl cl is connected to Gemini 2.5 Pro, the most advanced Google model for coding and simulated reasoning. It can create and modify code for you directly inside the terminal, but you can also call other Google models to generate images or videos without leaving the safety of your terminal coconut. It is essentially an ambient coding from the command line.
This tool is completely open source, so that developers can inspect the code and help improve it. The opening extends to the way you configure the AI agent. It supports the context protocol of the model (MCP) and the group extensions, allowing you to personalize your terminal as you see fit. You can even include your own System – Gemini CLI prompts is based on gemini.md files, which you can use to modify the model for different tasks or teams.
Now that Gemini 2.5 Pro is generally available, Gemini Code Assist has been upgraded to use the same technology as Gemini Cli. The Assist code fits into FDI such as the VS code for times when you need a richer environment. The new agent mode in the Assist code allows you to give the AI more general instructions, such as “adding the management of the dark mode to my application” or “Build my project and correct errors”.
