Instructions to use mikoy92/MiniMax-Music3-MLX with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mikoy92/MiniMax-Music3-MLX with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir MiniMax-Music3-MLX mikoy92/MiniMax-Music3-MLX
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
| license: other | |
| license_name: minimax-music3-community | |
| license_link: https://huggingface.co/MiniMaxAI/MiniMax-Music3 | |
| base_model: MiniMaxAI/MiniMax-Music3 | |
| library_name: mlx | |
| pipeline_tag: text-to-audio | |
| tags: | |
| - mlx | |
| - apple-silicon | |
| - music-generation | |
| - minimax | |
| # MiniMax Music 3 — MLX | |
| Converted [MiniMaxAI/MiniMax-Music3](https://huggingface.co/MiniMaxAI/MiniMax-Music3) weights for Apple Silicon via [mikolaj92/minimax-music3-mlx](https://github.com/mikolaj92/minimax-music3-mlx). | |
| This is a layout remap (PyTorch conv kernels → MLX, weight-norm fused, bf16). It is not a fine-tune. | |
| All variants: [MiniMax Music 3 MLX collection](https://huggingface.co/collections/mikoy92/minimax-music-3-mlx-6a7eb2a8e09d246e6d029fad) | |
| | Precision | Repo | Size | | |
| |---|---|---| | |
| | bf16 | [mikoy92/MiniMax-Music3-MLX](https://huggingface.co/mikoy92/MiniMax-Music3-MLX) | ~23.5 GB | | |
| | 8-bit | [mikoy92/MiniMax-Music3-MLX-8bit](https://huggingface.co/mikoy92/MiniMax-Music3-MLX-8bit) | ~12 GB | | |
| | 6-bit | [mikoy92/MiniMax-Music3-MLX-6bit](https://huggingface.co/mikoy92/MiniMax-Music3-MLX-6bit) | ~10 GB | | |
| | 4-bit | [mikoy92/MiniMax-Music3-MLX-4bit](https://huggingface.co/mikoy92/MiniMax-Music3-MLX-4bit) | ~7 GB | | |
| ## Use | |
| ```bash | |
| git clone https://github.com/mikolaj92/minimax-music3-mlx | |
| cd minimax-music3-mlx | |
| uv venv --python 3.12 .venv | |
| source .venv/bin/activate | |
| uv pip install -e . | |
| hf download mikoy92/MiniMax-Music3-MLX --local-dir ./weights/mlx | |
| python -m minimax_music3_mlx.generate \ | |
| --weights ./weights/mlx \ | |
| --lyrics $'[verse]\n...\n[chorus]\n...' \ | |
| --prompt "Genre: acoustic pop. BPM: 96. Warm female vocal." \ | |
| --duration 30 \ | |
| --steps 30 \ | |
| --output song.wav | |
| ``` | |
| ## License | |
| Weights remain under the [MiniMax-Music3 Community License](https://huggingface.co/MiniMaxAI/MiniMax-Music3) (Copyright 2026 MiniMax). Inference code in the GitHub repo is Apache-2.0. | |
| Commercial products that ship the model must display “MiniMax-Music3”. Yearly revenue above USD 20M needs a separate MiniMax authorization (`api@minimax.io`). | |