Save the voice note as a file, pick it here and read what was said. The audio is opened inside this browser tab and goes nowhere else. No account, no watermark.
Checking what your browser supports...
Hi, I'll be ten minutes late. Can you start without me?
Would you rather talk than type? StarWhisper lets you use the same Whisper in any text box on your PC or Mac. Press a shortcut, speak, and the words land in your email, document or chat. Free for 2,000 words a week, no account.
Try it freeWhatsApp keeps voice notes inside the chat, so the first job is to save one as a file. Pick the way that matches where you read your messages.
Menu names shift a little between WhatsApp versions. This page takes one file at a time, so for a batch you pick, wait, and click New file for the next.
Voice notes are short, so Base is usually enough. If a name or a place comes out wrong, that's the first spot to check by ear.
A voice note is often something you wouldn't want on a stranger's server. Here the file isn't uploaded. Your browser reads it and Whisper runs in the same tab. The only things that come in are the page, the reading library and the model files. You can watch the Network tab of your browser's developer tools to confirm that no file goes out.
The page does send two kinds of data, but neither includes your file or your words. If you accept cookies, our usual site analytics load as they do on every page. And the page sends a few anonymous counts to our server, without a cookie: when you pick a file (its rough size and how the page reads it), when a transcription finishes (the model, CPU or GPU, a rough length such as "1 to 5 minutes", and the speed), and when something fails (the step that failed). The code is open source on GitHub, so you can read exactly what it does.
The page also opens in your phone's browser. Browsers that can't read a file in pieces stop at 1 GB on a computer and 200 MB on a phone, and a voice note is far below both. Phones offer the Tiny and Base models only, which suits short notes. A computer handles long ones better. The page doesn't tell speakers apart. It writes what was said, without names.