# MLX

MLX: An array framework for Apple silicon

Repository: https://github.com/ml-explore/mlx
Canonical: https://ross.abutalabs.com/products/mlx
Homepage: https://ml-explore.github.io/mlx/
Language: C++
License: MIT
License Family: permissive
Topics: mlx
Last push: 2026-08-26T23:41:57+00:00
Link (homepage): https://ml-explore.github.io/mlx/

## Health v2 (maintenance only)
Score: 94/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 99, release rhythm 99, longevity 72
- inputs: {"age_days": 1009, "days_push": 7, "days_rel": 8, "gap_med": 13, "n_releases_24m": 44}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 28172, forks 2186 (observed 2026-08-28T04:11:48.152682+00:00)

## What it is
MLX is an array computation framework for machine learning on Apple silicon, developed by Apple ML research. It offers NumPy-like Python APIs plus C++ and Swift bindings, with autodiff, lazy evaluation, dynamic graphs, and a unified memory model across CPU and GPU.

## Use cases
- train and run machine learning models on Apple silicon
- run LLMs locally on a Mac
- finetune LLaMA with LoRA on Apple GPU
- generate images with Stable Diffusion on macOS
- run Whisper speech recognition locally
- NumPy-like array math with GPU acceleration on Mac

## When to choose
- you develop or deploy ML models on Apple silicon Macs
- you want PyTorch-like training APIs optimized for unified memory
- you want to run or finetune LLMs locally on macOS

## When to avoid
- you need NVIDIA CUDA GPUs or Linux/Windows servers
- you need the broad ecosystem and production tooling of PyTorch or JAX
- your project targets non-Apple hardware

## Facets
- artifact type: framework
- maturity: active
- function: machine-learning, deep-learning, llm-inference, llm-training, stable-diffusion, speech-recognition, gpu-computing
- domain: machine-learning, deep-learning, large-language-models, artificial-intelligence, gpu-computing
- platform: python, cpp
- tags: apple-silicon, numpy-like, autodiff, unified-memory, mlx, macos, swift, gpu

## Member repositories
- ml-explore/mlx (main) score 94
- ml-explore/mlx-examples (examples) score 61
- ml-explore/mlx-swift-examples (examples) score 73
- ml-explore/mlx-swift (sdk) score 91

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:11:48.152682+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T16:54:26.452135+00:00, confidence not recorded.
  - readme: https://github.com/ml-explore/mlx (fetched 2026-08-28T04:11:48.152682+00:00, sha fbb17a7f3ae0)
  - homepage: https://ml-explore.github.io/mlx/ (fetched 2026-08-29T07:51:02.409512+00:00, sha 44136fa355b3)
- Data as of 2026-08-30T08:39:29.467469+00:00.