comfy-mss

ComfyUI custom nodes for pymss, a Python music source separation package.

Nodes

• mss_separate: separates a ComfyUI AUDIO stream with non-VR pymss models.
• vr_separate: separates a ComfyUI AUDIO stream with VR/UVR pymss models.
• pymss MSS Params: optional parameter input for mss_separate.
• pymss VR Params: optional parameter input for vr_separate.
• pymss Model List: lists known pymss models and stem metadata as JSON.

Install pymss

From PyPI:

.\.venv\Scripts\python.exe -m pip install pymss

For local pymss development from E:\vs\pymss:

.\.venv\Scripts\python.exe -m pip install -e E:\vs\pymss

Restart ComfyUI after installing or changing the dependency.

Notes

• Use ComfyUI's built-in Load Audio and Save Audio / Preview Audio nodes for IO.
• mss_separate and vr_separate outputs are refreshed by the frontend extension after choosing a model.
• The backend declares up to 64 audio stems, which covers the current pymss catalog maximum of 53 stems.
• Model files default to ComfyUI/models/pymss. The folder is created automatically when the node loads.
• The default model folder can be changed at runtime with COMFY_MSS_MODEL_DIR. PYMSS_MODEL_DIR is also supported for pymss compatibility.
• The node-level model_dir input overrides both environment variables for that execution.
• Both parameter nodes include extra_params_json for advanced pymss inference overrides, for example:

{
  "batch_size": 2,
  "chunk_size": 352800
}