Show HN: Mlx-serve – LLM inference server for Apple Silicon, written in Zig

Category: infrastructure

Tags: llm-inference, apple-silicon, zig

Score: 5.3/10 (Innovation: 5, Technical: 6, Documentation: 2, Utility: 5)

Mlx-serve is an LLM inference server specifically optimized for Apple Silicon, written in the Zig programming language. It targets users running large language models locally on Mac hardware, combining a niche use case with a modern systems language. The project is interesting for leveraging Apple's MLX framework via a lightweight server, but lacks comprehensive documentation.

Target audience: machine learning engineers, backend devs

Repository: https://mlxserve.com/

View on Hacker News