ModelMesh — Native macOS Multi-Node llama.cpp & GGUF Orchestrator

ModelMesh is a macOS multi-node llama.cpp and GGUF orchestrator that pools unified memory across multiple networked Macs to run models exceeding a single machine's RAM. It uses direct Objective-C++ bindings to llama.cpp with a native AppKit UI and provides OpenAI-compatible API endpoints for integration into developer tools and custom scripts. All prompts, model weights, and inference traffic remain local to the user's network. Open article →