openunmix
PyTorch-based music source separation toolkit
Decision gist · record as of 2026-08-14
Yes, if you need music source separation and accept the non-commercial license on the default model. The package is stable, low-friction to install, and provides production-ready models with no known vulnerabilities. However, dormancy (850 days since last release) means no active maintenance or bug fixes; consider it if your use case matches the pre-trained models and you do not require ongoing support.AI-flagged interpretation of the facts on this page — verify before relying
Before you install
- Requires PyTorch 1.8+ and torchaudio; GPU acceleration optional but recommended for performance.
- Default model (`umxl`) weights are non-commercial use only.
- Low friction installation with a pure Python wheel.
License · maintenance · safety
MIT (permissive) — MIT license (permissive) applies to the package code. Note that the default `umxl` model weights are licensed CC BY-NC-SA 4.0 for non-commercial use only, which may restrict deployment in commercial applications.
last release 2024-04-16 (850 days) · last repo commit 2024-06-17 · 1,501 stars
0 known vulnerabilities (OSV.dev, 2026-08-14) · 302,651 downloads/mo, #7,819 on PyPI
Alternatives
Verify before relying
pip install openunmix
import openunmix
separator = openunmix.umxl()
# Load audio and separate
waveform, sr = torchaudio.load('track.wav')
stems = separator(waveform)- Whether pre-trained model weights are automatically downloaded on first use or must be manually fetched.
- Memory and compute requirements for real-time or batch separation on typical hardware.
- Whether the package supports GPU acceleration beyond what torch and torchaudio provide.
What it is and what it does
Open-Unmix is a PyTorch-based toolkit for music source separation that decomposes mixed audio into individual instrumental stems using pre-trained deep neural networks. It provides ready-to-use models trained on the MUSDB18 dataset to extract vocals, drums, bass, and other instruments from pop music, plus a speech enhancement model. The core architecture is a three-layer bidirectional LSTM that learns to predict magnitude spectrograms of target sources; the `Separator` class combines multiple source models and applies a differentiable Wiener filter to produce final waveforms.
The package is designed for researchers, audio engineers, and artists who need to isolate or analyze individual instruments in recordings. It operates on waveforms or pre-computed spectrograms, handles arbitrary audio lengths due to its recurrent architecture, and includes a command-line interface for batch processing. Installation is straightforward via pip, though the default model weights carry a non-commercial license restriction.
Use it for
- Extract vocal tracks from songs for remixing, karaoke, or vocal analysis without manual editing.
- Isolate drum patterns or bass lines for music production, sampling, or rhythm study.
- Separate speech from background noise using the `umxse` speech enhancement model.
- Batch process music libraries to create stem versions for archival or downstream analysis.
- Research music information retrieval or develop custom source separation models using the provided architecture.
Worth the install?
AI-flagged interpretation of the facts on this page. Verify before relying on it.
Yes, if you need music source separation and accept the non-commercial license on the default model.
The package is stable, low-friction to install, and provides production-ready models with no known vulnerabilities. However, dormancy (850 days since last release) means no active maintenance or bug fixes; consider it if your use case matches the pre-trained models and you do not require ongoing support.
Install
openunmix on PyPI
Before you install
Low friction installation with a pure Python wheel. The package is dormant (last release 850 days ago) but stable and Production/Stable classified; no active maintenance signals, though the repository remains unarchived with 1501 stars.
Requires PyTorch 1.8+ and torchaudio; GPU acceleration optional but recommended for performance. Default model (`umxl`) weights are non-commercial use only.
License in practice
MIT license (permissive) applies to the package code. Note that the default `umxl` model weights are licensed CC BY-NC-SA 4.0 for non-commercial use only, which may restrict deployment in commercial applications.
Quickstart
pip install openunmix
import openunmix
separator = openunmix.umxl()
# Load audio and separate
waveform, sr = torchaudio.load('track.wav')
stems = separator(waveform)Verify before relying
- Whether pre-trained model weights are automatically downloaded on first use or must be manually fetched.
- Memory and compute requirements for real-time or batch separation on typical hardware.
- Whether the package supports GPU acceleration beyond what torch and torchaudio provide.
Package facts
| License | MIT permissive |
| Python support | Supports the current Python release >=3.9 |
| Install friction | Low. Pure-Python wheel |
| Runtime dependencies | 4 packagesnumpytorchaudiotorchtqdm |
| Maintenance | Dormant 850 days since the last release |
| Last repo commit | |
| First released | |
| Downloads | 302,651 / month, #7,819 on PyPI 30-day window, as of 2026-08-14 |
| Known vulnerabilities | None known OSV.dev, checked 2026-08-14 |
| Classifiers | Development Status :: 5 - Production/StableEnvironment :: ConsoleIntended Audience :: DevelopersLicense :: OSI Approved :: MIT LicenseOperating System :: OS IndependentProgramming Language :: PythonProgramming Language :: Python :: 3 :: OnlyProgramming Language :: Python :: 3.10Programming Language :: Python :: 3.11Programming Language :: Python :: 3.12Programming Language :: Python :: 3.8Programming Language :: Python :: 3.9Topic :: Software Development :: Libraries :: Python ModulesTopic :: Software Development :: Quality Assurance |
Evidence: openunmix-1.3.0-py3-none-any.whl
Tags
Let your AI agent find packages like this
Example. Real query, live index.
You found this page by searching. An agent finds it by wishing: SkillFed indexes 14,416 PyPI packages by what they can do, searchable in plain language.
wish › “music source separation”
- openunmixSeparates music into individual stems (vocals, drums, bass, other…
- demucsDemucs separates music into individual stems—drums, bass, vocals, and…
- audio-separatorSeparates audio files into multiple stems (vocals, instruments,…
Give your agent the search over MCP, or paste the wish link into any chat.
More Python Modules packages
Converts domain names between Unicode and ASCII-compatible encoding (Punycode) according to IDNA 2008 and Unicode Technical Standard 46, with security validation and broader script coverage than the standard library.
Install it if you work with internationalized domain names, need to validate domains, or use HTTP clients that depend on it transitively.
Setuptools is a Python build backend and package management tool that handles building, distributing, and installing Python packages, including support for C/C++ extension modules.
PyYAML parses and emits YAML 1.1 data format, enabling serialization and deserialization of configuration files and Python objects to and from human-readable YAML text.
Pydantic validates Python data structures against type hints, coercing and checking input at runtime to ensure it matches a declared schema.
Provides reusable metadata objects for use with PEP-593 `typing.Annotated` to express common constraints like bounds, collection sizes, and predicates on types.
Install it if you use or build libraries that need to express type constraints in a standardized, inspectable way—or if you want to annotate your own types with…
Provides runtime tools to inspect and introspect Python type annotations, enabling programmatic examination of type hints at execution time.
See also asteroid-filterbanks · audio-separator · demucs · torchaudio · laion-clap · torchlibrosa · encodec · snac · descript-audio-codec