mlx-lm
brew install mlx-lmRun LLMs with MLX
https://github.com/ml-explore/mlx-lm
License: MIT
Development: Pull requests
Formula JSON API: /api/formula/mlx-lm.json
Formula code: mlx-lm.rb on GitHub
Bottle (binary package) installation support provided for:
| macOS on Apple Silicon |
tahoe | ✅ |
|---|---|---|
| sequoia | ✅ | |
| sonoma | ✅ |
Current versions:
| stable | ✅ | 0.31.3 |
| head | ⚡️ | HEAD |
Revision: 3
Depends on:
| certifi | 2026.7.22 | Mozilla CA bundle for Python |
| libyaml | 0.2.5 | YAML Parser |
| mlx | 0.32.1 | Array framework for Apple silicon |
| numpy | 2.5.3 | Package for scientific computing with Python |
| protobuf | 36.1 | Protocol buffers (Google's data interchange format) |
| python@3.14 | 3.14.7 | Interpreted, interactive, object-oriented programming language |
| sentencepiece | 0.2.2 | Unsupervised text tokenizer and detokenizer |
Depends on when building from source:
| cmake | 4.4.3 | Cross-platform make |
| coreutils | 9.11 | GNU File, Shell, and Text utilities |
| pkgconf | 3.0.7 | Package compiler and linker metadata toolkit |
| rust | 1.98.1 | Safe, concurrent, practical language |
Requires: macOS >= 13
Binaries:
mlx_lm,
mlx_lm.awq,
mlx_lm.benchmark,
mlx_lm.cache_prompt,
mlx_lm.chat,
mlx_lm.convert,
mlx_lm.dwq,
mlx_lm.dynamic_quant,
mlx_lm.evaluate,
mlx_lm.fuse,
mlx_lm.generate,
mlx_lm.gptq,
mlx_lm.lora,
mlx_lm.manage,
mlx_lm.perplexity,
mlx_lm.server,
mlx_lm.share,
mlx_lm.upload
Service:
| Start now and at login: | brew services start mlx-lm |
| Run directly: | $HOMEBREW_PREFIX/opt/mlx-lm/bin/mlx_lm.server |
Analytics:
| 30 days | 90 days | 365 days | |
|---|---|---|---|
| Installs | 411 | 1,805 | 6,444 |
Installs (--HEAD) | 1 | 9 | 35 |
| Installs on Request | 410 | 1,803 | 6,442 |
Installs on Request (--HEAD) | 1 | 9 | 35 |
Build Errors (--HEAD) | 3 |