SoulX-Singer-8bit

SoulX-Singer-8bit

Apple MLX safetensors checkpoint for Soul-AILab/SoulX-Singer, including both SoulX-Singer SVS and SoulX-Singer-SVC weights.

Collection: https://huggingface.co/collections/mlx-community/soulx-singer-mlx-6a1c0525d0911ea400102840

This is not a pure end-to-end MLX audio runtime yet. The weights are converted to an MLX-friendly safetensors layout, while full audio generation currently uses the official PyTorch model structure through ailuntx/SoulX-Singer-MLX. This checkpoint uses MLX affine 8bit quantization; the bridge dequantizes weights before loading the official PyTorch modules.

TL;DR

Quick Start

git clone https://github.com/ailuntx/SoulX-Singer-MLX.git
cd SoulX-Singer-MLX
conda create -n soulxsinger-mlx -y python=3.10
conda activate soulxsinger-mlx
python -m pip install -U pip
python -m pip install -r requirements.txt mlx safetensors huggingface_hub hf_transfer

HF_HUB_ENABLE_HF_TRANSFER=1 hf download mlx-community/SoulX-Singer-8bit --local-dir ./models/SoulX-Singer-8bit
hf download openai/whisper-base --local-dir pretrained_models/openai__whisper-base

Run a short SVS bridge test:

PYTORCH_ENABLE_MPS_FALLBACK=1 \
SOULX_WHISPER_MODEL=pretrained_models/openai__whisper-base \
python scripts/inference_mlx_bridge.py \
  --model ./models/SoulX-Singer-8bit \
  --component svs \
  --device mps \
  --prompt_wav_path example/audio/zh_prompt.mp3 \
  --prompt_metadata_path example/audio/zh_prompt.json \
  --target_metadata_path example/audio/zh_target.json \
  --control melody \
  --n_steps 1 \
  --cfg 1 \
  --save_dir outputs_mlx_bridge/svs

Run an SVC bridge test:

PYTORCH_ENABLE_MPS_FALLBACK=1 \
SOULX_WHISPER_MODEL=pretrained_models/openai__whisper-base \
python scripts/inference_mlx_bridge.py \
  --model ./models/SoulX-Singer-8bit \
  --component svc \
  --device mps \
  --prompt_wav_path example/audio/zh_prompt.mp3 \
  --target_wav_path example/audio/music.mp3 \
  --prompt_f0_path example/audio/zh_prompt_f0.npy \
  --target_f0_path example/audio/music_f0.npy \
  --n_steps 1 \
  --cfg 1 \
  --save_dir outputs_mlx_bridge/svc

Variants

Layout

SoulX-Singer-8bit/
|-- config.json
|-- config.yaml
|-- mlx_manifest.json
|-- svs/
|   |-- model.safetensors.index.json
|   `-- model-00001-of-000xx.safetensors
`-- svc/
    |-- model.safetensors.index.json
    `-- model-00001-of-000xx.safetensors

This repo has 6 SVS shard(s) and 6 SVC shard(s).

Validation

Local Apple Silicon validation:

For quantized checkpoints, the bridge loader dequantizes MLX affine tensors into the official PyTorch module shapes for compatibility testing. Native all-MLX inference is planned as a later runtime step.

License

The converted weights follow the upstream SoulX-Singer Apache-2.0 release.

Citation

@misc{soulx-singer-mlx,
  title  = {SoulX-Singer-MLX: Apple MLX safetensors port of SoulX-Singer},
  author = {ailuntx},
  year   = {2026},
  url    = {https://github.com/ailuntx/SoulX-Singer-MLX},
}

@misc{soulxsinger,
  title={SoulX-Singer: Towards High-Quality Zero-Shot Singing Voice Synthesis},
  author={Jiale Qian and Hao Meng and Tian Zheng and Pengcheng Zhu and Haopeng Lin and Yuhang Dai and Hanke Xie and Wenxiao Cao and Ruixuan Shang and Jun Wu and Hongmei Liu and Hanlin Wen and Jian Zhao and Zhonglin Jiang and Yong Chen and Shunshun Yin and Ming Tao and Jianguo Wei and Lei Xie and Xinsheng Wang},
  year={2026},
  eprint={2602.07803},
  archivePrefix={arXiv},
  primaryClass={eess.AS},
  url={https://arxiv.org/abs/2602.07803},
}