Ultimate Vocal Remover
github.com
- Free plan
- Windows
- macOS
- Linux
What Ultimate Vocal Remover is
Ultimate Vocal Remover is a free desktop application for separating audio sources and removing vocals from files. It uses deep neural network source-separation models and supports batch processing. Exports can be WAV, FLAC, or MP3, with stem choices including vocals, instrumental, other, bass, drums, guitar, piano, synthesizer, strings, woodwinds, brass, and wind instruments. The project describes itself as fully open source and free to use and modify. Its GUI code uses the MIT license, while developers using the models are asked to credit UVR and its developers. UVR runs on Linux, macOS, and Windows. Windows installation targets Windows 10 or higher and should use the main C: drive. The bundled macOS release targets Big Sur and later, with Apple Silicon and Intel downloads. Linux instructions cover Debian- and Arch-based systems. FFmpeg is needed for non-WAV processing, and Rubber Band supports time-stretch and pitch-shift tools.
Who it suits
It suits musicians, editors, and other users who need vocal or stem separation on a desktop. It is also relevant to developers seeking open-source software they can modify.
Good to know, and what to watch
- Free plan with no charge
- Batch processing supported
- Exports WAV, FLAC, and MP3
- Runs on Linux, macOS, and Windows
- Many selectable stem types
- GPU conversion needs an Nvidia GTX 1060 6GB
- AMD Radeon support is limited
- Models are computationally intensive
- Windows installation requires the main C: drive
Plans
Every plan and priceUltimate Vocal Remover GUI
Free plan
The verdict
Ultimate Vocal Remover combines broad stem options, batch processing, and desktop support across three operating systems. Hardware requirements and limited AMD support matter when planning GPU-based conversions.
What the maker publishes
- AMD support
- AMD Radeon GPU support is limited, though the project links to a working branch for AMD GPU users. source
- Dependencies
- UVR relies on FFmpeg to process non-WAV files and Rubber Band for time-stretch and pitch-shift tools. source
- GPU requirements
- The project lists an Nvidia GTX 1060 with 6GB as the minimum for GPU conversions and recommends Nvidia GPUs with at least 8GB of VRAM. source
- License
- The UVR GUI code is MIT-licensed, and developers using its models are asked to credit UVR and its developers. source
- Linux
- The repository provides manual installation instructions for Debian- and Arch-based Linux systems. source
- macOS
- The bundled macOS version targets Big Sur and above and has separate downloads for Apple Silicon and Intel Macs. source
- Model training
- UVR's core developers trained the models included in the package, except the Demucs v3 and v4 4-stem models. source
- Open source
- The project says it is 100% open source and free for anyone to use and modify. source
- Performance
- The project says conversion times depend significantly on hardware and that its models are computationally intensive. source
- Purpose
- The application uses deep neural network source separation models to remove vocals from audio files. source
- Security note
- The macOS opening instructions advise that users may need to disable Gatekeeper checks and remove the app's quarantine attribute to bypass notarization. source
- Support
- Users are directed to GitHub Issues for help and asked to include detailed reports, with error logs when possible. source
- Windows
- The Windows installer targets Windows 10 or higher, and installation to the main C: drive is required to avoid instability. source
Checked against the maker's pages on 3 Oct 2026.