MLX: An array framework for Apple silicon

Fields

id

sources/src5-mlx-github

URL

https://github.com/ml-explore/mlx

Publisher

Apple Machine Learning Research

Kind

repo

Published

2023-12-01

Accessed

2026-10-04

notes

MLX is an array framework for ML on Apple silicon. Key features: NumPy-like Python API with C++/C/Swift APIs, composable function transformations (autodiff, vectorization, graph optimization), lazy computation, dynamic graph construction, multi-device (CPU+GPU), unified memory model. Inspired by NumPy, PyTorch, JAX, ArrayFire. 28.6k stars. The unified memory model eliminates data transfer between devices.

Notes

MLX is an array framework for ML on Apple silicon. Key features: NumPy-like Python API with C++/C/Swift APIs, composable function transformations (autodiff, vectorization, graph optimization), lazy computation, dynamic graph construction, multi-device (CPU+GPU), unified memory model. Inspired by NumPy, PyTorch, JAX, ArrayFire. 28.6k stars. The unified memory model eliminates data transfer between devices.

Cross-references

This record references no other records.