Audio Profanity Remover — Clean Version Maker

By default it mutes Russian obscene roots and their derivatives. Rude words and mild euphemisms can be enabled with separate toggles, and you can add any words to the custom blacklist.

No — it handles any audio the same way: songs, podcasts, voice messages, lectures, interviews, audiobooks. You can also upload a video file and its soundtrack will be processed.

Yes — the model is trained on real-world recordings and is robust to background music, though very dense mixes can blur individual words. Every detection is visible in the transcript, and a missed spot can always be muted manually by selecting it on the waveform.

Yes: pick the "My own sound" mode and upload a short audio file — it will replace every muted section. The first 30 seconds of the file are used, looping over longer sections, with adjustable volume.

Currently the recognizer is tuned for Russian speech (GigaAM, trained on 700,000 hours of Russian audio). More languages are in development — and manual muting works for recordings in any language.

No — words are muted in place, so duration and sync are fully preserved. Smooth fades at mute edges prevent clicks.