LlamaBarn
A tiny menu bar app that makes running local LLMs as easy as connecting to Wi-Fi
Free LLM Server Open Source
Overview
LlamaBarn is a lightweight native macOS menu bar application that simplifies running local large language models. Built with Swift and weighing only 12 MB, it provides a pre-configured model catalog, automatic optimization for your Mac's specs, and a standard OpenAI-compatible API at localhost:2276 for seamless integration with chat apps, editors, and custom scripts.
Architecture: Apple Silicon
Key Features
- Native macOS app built with Swift - only 12 MB
- Pre-configured model catalog optimized for your Mac's specs
- OpenAI-compatible API at localhost:2276
- Built-in web interface for direct chat
- Works with chat applications, editors, CLI tools, and custom scripts
- Automatic handling of quantization, context windows, and device settings
- Installable via Homebrew or direct download
