BitRun

BitRun is a cross-platform command-line tool that allows running and managing ultra-efficient 1-bit and 1.58-bit ternary large language models (LLMs) such as Microsoft's BitNet b1.58 on macOS, Windows, and Linux CPUs without requiring compilation. It supports pulling models from Hugging Face, running interactive chats, creating custom personas, and serving an OpenAI-compatible REST API, optimized for CPU inference with low energy consumption and no GPU required.