infrctl is a simple command-line tool for running private AI locally. Instead of forcing users to sort through hundreds of confusing model names, infrctl gives them one clean CLI to run supported open model families like Qwen, DeepSeek, Llama, Gemma, and Phi on their own machine.

Users can install models, chat locally, ask one-off questions, check hardware compatibility, and expose a local API for apps, agents, scripts, or developer tools.

infrctl is built for people who want AI without cloud APIs, API keys, hosted dashboards, or complicated setup.

$infrctl is built on Base around private local AI infrastructure.