llamafile

Ship an entire LLM as one cross-platform executable file.

llamafile, from Mozilla Ocho, packages an LLM and its weights into a single executable that runs on macOS, Windows, Linux, and BSD without installation or dependencies. It combines llama.cpp with Cosmopolitan Libc to produce one portable file that boots a local chat UI and an OpenAI-compatible server. It is designed to make distributing and running local models as simple as downloading and running one file.