LALAL.AI vs Open-Unmix

LALAL.AI uses AI to separate vocals and instrument stems, clean voice recordings, change or clone voices, and reduce echo and reverb. Its available stem types include vocals, instrumental, drums, bass, electric and acoustic guitar, piano, synthesizer, strings, and wind instruments. The web vocal remover accepts audio and video formats, including MP3, WAV, FLAC, MP4, and MOV. Fast Queue prioritizes processing, while Relaxed Queue processes as server capacity allows; both provide the same separation quality. Usage minutes are calculated by multiplying file length by the number of selected stem types. The desktop app supports Windows, macOS, and Ubuntu Linux, with batch processing for up to 20 files; Pro subscribers can optionally process locally offline. The API and VST plugin offer other ways to use the tools. The free Starter plan includes 10 Relaxed Queue minutes, but does not allow full downloads. Lite costs $7.50 a month billed annually.

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.