Remove background noise and recover clean speech from any recording, right in your browser. Powered by machine learning compiled to WebAssembly. Free to use, no account needed. Air-Gap privacy guarantee - Runs on your device, no cloud server or databases.
Outdoors interviews, meetings, calls, lectures and field recordings where speech is the main subject and the surroundings are noise: traffic, fans, crowds, hum, room reverb.
Music streams such as songs or rap where the instrumental is dominant. The model treats music as noise, so expect heavy artefacts or a muffled vocal. It is a speech enhancer, not a stem separator.