RVZWN//TOOLS

> SECTOR_03 // AUDIO & VIDEO // SPEECH_TO_TEXT

Speech to Text

Transcribe speech from any audio or video file with the Whisper model running on your device. Choose the language or let it detect it, get the text with timestamps, edit it, and save as TXT, SRT or VTT subtitles. The sound never leaves your computer.

LOCAL_PROCESSING // NO_UPLOAD

How it works
  1. Open a recording or a video (its sound is used), up to 30 minutes. It is decoded on your device.
  2. Choose the spoken language, or leave it on automatic. You can also have everything translated into English.
  3. Press Transcribe. The speech model (Whisper, small size) runs inside this page; the first time it loads about 250 MB from this site and your browser keeps it afterwards.
  4. The text comes back in lines with their start times. Click a line to correct it, then copy the text or save it as TXT, or as SRT / VTT subtitles you can use in a video player or editor.
It runs on your processor, so a long recording takes a few times its length; Cancel stops it at once. Clear speech in a quiet room works best, and names, numbers and rare words deserve a second look. Nothing is uploaded: the sound and the text stay on your device.