open audio
Built for the community, with the community.
Free • open source • no account

Clean audio online for free.
Open-source noise reduction and subtitles.

Use a free, open-source audio cleaner to reduce background noise in speech or create SRT, VTT and text transcripts. No account, ads or paid API.

Free & open sourceNo accountFiles are deleted after processing
Choose an audio or video file
MP3, WAV, M4A, OGG, OGA, OPUS, FLAC, AAC, WMA, MP4, WebM, MOV or MKV • Up to 32 MB / 5 min

Reduce background noise

Bring speech forward and reduce steady noise. Download a cleaned WAV file.

Transcribe speech

Auto-detect or choose the spoken language. Download subtitles as SRT, VTT or plain text.

Your recording stays yours.The file is uploaded briefly for processing and deleted when the download finishes. No account; we do not keep recordings or transcripts.
Open source

See how the tools work, contribute, or run them on your own server.

No paid API

Tools run open-source models on our own server.

Fair shared access

One conversion runs at a time so everyone can use it for free. Files are limited to 5 minutes.

Free, open-source online audio cleaner and subtitles

Open Audio is a free, open-source online audio cleaner for speech recordings. Reduce background noise with DeepFilterNet or transcribe speech with Whisper.cpp, then download WAV, SRT, VTT or text. The source code is public on GitHub; no account, ads or paid API are required.

Frequently asked questions

Is Open Audio really free?

Yes. It is free and open source, with no account, subscription, ads or paid API. One conversion runs at a time to share the community server fairly.

Which file types can I upload?

MP3, WAV, M4A, OGG, OGA, OPUS, FLAC, AAC, WMA, MP4, WebM, MOV and MKV. Files are limited to 32 MB and 5 minutes.

Are my audio files stored on the server?

No. Processing uses a temporary folder; input and output files are removed after the download response finishes. See the privacy page.

Which subtitle formats are available?

Download speech transcription as SRT, WebVTT (VTT) or plain text. Results vary with language, microphone and recording quality.