Privacy

Your audio, video and transcripts stay on your device. Wavetotext has no accounts, no ads and no analytics.

What is downloaded

When you first transcribe, your browser downloads the Transformers.js library from jsDelivr and the Whisper model files from Hugging Face. Speaker labels also download the pyannote and WeSpeaker models from Hugging Face. Videos your browser can't open download the ffmpeg converter from jsDelivr. These services see your IP address, as with any download. Nothing about your audio is sent to them.

What is stored

Finished transcripts are saved in your browser's storage, listed under Saved on this device. The audio is not saved. Model files are kept in your browser's cache so they don't download again. Your theme, quality and graphics card choices are saved in local storage. Clear this site's data in your browser settings to remove them.

Microphone

If you press Record, your browser asks for microphone access. The recording stays in the page's memory and is gone when you close the tab.

Other services

Fonts load from Google Fonts, which receives your IP address when it sends them. Our web host may keep standard server logs, such as IP address and pages visited, for security.