Open-Unmix vs UnMixLabs

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.

UnMixLabs is a web service and API that separates finished audio into vocal and instrumental stems. It accepts MP3, WAV, FLAC, M4A, OGG, MP4, and MKV files, with 4-stem, 6-stem, 53-stem, and vocal-isolation modes. Stem types include vocals, drums, bass, guitar, piano, other instruments, and instrumental. Each job includes BPM, musical-key, and chord-progression analysis. Optional processing offers vocal dereverb, stem-to-MIDI, beat-grid export, synced lyrics, karaoke video, and audio cleanup presets. Outputs can include WAV, MP3, FLAC, MIDI, beat-grid TSV, REAPER sessions, lyric SRT files, and karaoke MP4s. Batch processing is available, and Studio lists API access. The Free plan costs nothing, allows 10 processing minutes monthly, one 100 MB file at a time, 4-stem splits, and 24-hour retention. Pro is $10.99 a month; Studio is $24.99 a month. The pricing page says service is free-only during beta.