07 / 25 · TOOL · Apple silicon

MLX Sharding

Pipeline-parallel local LLM inference across multiple Macs, including shard servers and an OpenAI-compatible API.

distributed inferencepipeline parallelismMLXOpenAI API

Best way to startInstall with pip

$ pip install mlx-sharding

macOS on Apple silicon

Version0.1.1

Updated2024-07-27

PriceFree

ProcessingLocal

What it is

Built for useful AI without a remote API.

MLX Sharding is part of the local-ai.cc catalog of released applications, command-line tools, and developer packages for Apple silicon. Its primary workflow runs on the Apple device described above.

Use the verified installation method on this page, then consult the project source for complete model requirements, examples, licenses, and troubleshooting.