🔒 The Whisper AI runs in your browser: the audio never leaves your device

🎙️ Turn audio into text, offline

WhatsApp voice note, meeting, class, interview: drop the file or record yourself, and get the text with timestamps, plus a subtitle file.

The model is downloaded once, then kept by the browser. One minute of audio takes about 10 to 60 seconds depending on the device.

Model: OpenAI's Whisper (MIT license), run with Transformers.js.