Hi Product Hunt! I'm Eralp, and I built AudioRinse.

Most online audio tools want your file first: upload it, wait in a queue, and hope it isn't kept. That's awkward for an interview, an unreleased song or a client's voice-over. So AudioRinse does the work in your browser tab instead. Open models (DeepFilterNet for noise, Whisper for transcripts, HTDemucs and Mel-RoFormer for stems, and a few more) download once and run on your own computer, some on the graphics chip through WebGPU. Your file never leaves the device: open the network tab while it runs and check.

There are 60 tools so far:

• Clean a recording: background noise, echo, hum, harsh S's, mouth clicks, clipping, loudness

• Songs: vocal remover, 4-stem splitter, BPM and key finder, key, pitch and speed changer

• Transcripts and subtitles with speaker labels, for Word, SRT, VTT, NVivo or MAXQDA

• Piano recording to sheet music (PDF, MusicXML, MIDI)

• A voice recorder and free converters (video to MP3, WAV to MP3 and more)

Every tool runs and previews free, without an account. The free plan includes 3 downloads a month; the converters and the recorder have no limit. Pro is $7.99 a month or $39.99 a year for unlimited downloads, files up to an hour, and video.

I'd love to know which tool you tried and where it fell short. Your browser and computer help too: making these models run well on every graphics chip is the hardest part to test.