Skip to content
SPECIFICTION
Get listed

Open-Unmix

github.com

Free plan WindowsmacOSLinuxSelf-hosted
Visit github.com
Spec score 7.8 #12 of 21 in Stem Separation Software

What it is

Open-Unmix is a PyTorch reference implementation for music source separation, aimed at researchers, audio engineers, and artists. Its ready-to-use models split pop music into vocals, drums, bass, and other instruments; a separate model, umxse, is for speech enhancement and was trained on the 28-speaker Voicebank+DEMAND corpus. Pre-trained choices include umxl, umxhq, umx, and umxse, with spectrogram variants also available. Each source model uses a three-layer bidirectional LSTM, a design the project says cannot run online or in real time. The code supports reproducing results, developing architectures, and adding data, while training is documented separately from the package. Install through pip or use the Docker container; pre-trained models can be loaded in Python or through torch.hub, and a Colab notebook supports online experimentation without local setup. The command line handles WAV, FLAC, and OGG, not MP3; stempeg can extend formats, including STEMS export. The software is MIT-licensed; umxl weights are non-commercial under CC BY-NC-SA 4.0.

Spec sheet

Purpose
Open-Unmix is a PyTorch reference implementation for music source separation aimed at researchers, audio engineers, and artists.Read 8 Oct 2026 · github.com
Integrations
Pre-trained models can be loaded from Python or through torch.hub, and a Google Colab notebook offers online experimentation without local setup.Read 8 Oct 2026 · github.com
Support
The project directs users to GitHub issues, Gitter chat, and Google Groups for technical help and support.Read 8 Oct 2026 · sigsep.github.io
Maker
The SigSep GitHub organization describes itself as providing open resources for audio source separation.Read 8 Oct 2026 · github.com
Audio formats
The command line supports WAV, FLAC, and OGG files but not MP3; stempeg can expand supported input and output formats, including STEMS export.Read 8 Oct 2026 · github.com
Installation
The Python package can be installed with pip, and the project also provides a Docker container.Read 8 Oct 2026 · github.com
Model architecture
Each source model uses a three-layer bidirectional LSTM, and the project states that this design cannot run online or in real time.Read 8 Oct 2026 · github.com
Model options
Pre-trained options include umxl, umxhq, umx, and umxse, with spectrogram variants also available.Read 8 Oct 2026 · github.com
Model weights license
The umxl weights are licensed for non-commercial use under CC BY-NC-SA 4.0.Read 8 Oct 2026 · github.com
Music stems
Its ready-to-use models separate pop music into vocals, drums, bass, and other instruments.Read 8 Oct 2026 · github.com
Security and compliance
The opened project pages state an MIT license but do not describe security controls or compliance certifications.Read 8 Oct 2026 · github.com
Software license
The repository lists the software license as MIT.Read 8 Oct 2026 · github.com
Speech enhancement
The project provides an umxse speech enhancement model trained on the 28-speaker Voicebank+DEMAND corpus.Read 8 Oct 2026 · github.com
Training and extensibility
The code is designed to let researchers reproduce results, develop architectures, and add their own data, while model training is documented separately from the package.Read 8 Oct 2026 · github.com

Plans and prices

Open-Unmix Free MIT-licensed software · umxl model weights are licensed for non-commercial use Read 8 Oct 2026 · github.com

Every Open-Unmix plan and price →

Strong points and limits

Strong points

  • Separates vocals, drums, bass, and other instruments
  • Includes a speech enhancement model
  • Supports model loading through Python and torch.hub
  • Free, MIT-licensed software

Limits

  • Model design cannot run online or in real time
  • Command line does not support MP3
  • umxl weights are licensed for non-commercial use

Who it suits

Open-Unmix suits researchers, audio engineers, and artists who want to separate music or explore and extend source-separation models. Its stated model design is not for online or real-time processing.

Verdict

Open-Unmix offers several pre-trained models and code for research and extension, with both local installation and Colab experimentation options. Check the audio format limits and the separate non-commercial license for umxl weights before using it.

Where Specifiction places it

Alternatives to Open-UnmixAll alternatives →