MLX-Tune wraps Apple's MLX in an Unsloth-style Python API (FastLanguageModel, SFTTrainer), so a training script written on a Mac can run on an NVIDIA GPU with Unsloth by changing the imports. It supports SFT, DPO and GRPO, and fine-tuning of vision, text-to-speech, speech-to-text, embedding and OCR models.
Trained models export to Hugging Face format or GGUF for Ollama and llama.cpp. It is an unofficial project, not made by Unsloth, and was previously called unsloth-mlx.
It needs an Apple silicon Mac on macOS 13.0 or later with 8 GB of memory (16 GB recommended) and Python 3.9+. It's free and open source (Apache 2.0); training runs on your Mac.
How to Get It
uv pip install mlx-tunepip install mlx-tuneOfficial downloads only. MLX never re-hosts an app, and never runs a command for you.
MLX Score
Ranked against the developer tools listings, from real public data only. How the score works.
Pairs Well With
Information
- Developer
- ARahim3
- Category
- Developer tools
- Price
- Free
- Your data
- Runs on your Mac
- Licence
- Apache-2.0
- Mac
- macOS 13.0 or later
- iPhone and iPad
- —
- Needs
- 8 GB memory, M1 or later
- Facts checked
- 2026-11-03
Sources
You Might Also Like
uvDev tools
TransformersDev tools
LiteLLMDev tools
Hugging Face CLI (hf)Dev tools
llmfitDev tools
Foundation Models SDK for PythonDev toolsMLX-Tune is a product of ARahim3; MLX isn’t affiliated with it. Something wrong? Write to support@mlx.app.