The AI Demuxer turns any song into a practice room.

Pick any audio file from the AI Demux import menu and openDAW splits it into four stems, drums, bass, other, and vocals. Each stem lands on its own track, so you can mute the guitar and play along, solo the drums and practice timing, or pull out the vocals for a quick study session.

Everything runs locally in the browser. The model is downloaded once and cached, and inference runs on WebGPU when available, or falls back to a Web Worker on the CPU.

Try it at https://opendaw.studio