Upload an audio file, get an accurate AI transcript in seconds. MP3, M4A, WAV, OGG, FLAC and more — up to 25 MB. No signup.
The Typally desktop app types your voice straight into any application — hold a key, talk, done. 7-day free trial.
This tool runs Whisper large-v3 — the same state-of-the-art speech recognition model used by the Typally desktop app. It handles accents, background noise, and technical vocabulary far better than the speech recognition built into your phone or browser.
Yes. Your file is sent over an encrypted connection, transcribed, and immediately discarded. We don't store your audio or your transcript — the text exists only in your browser.
The free tool allows 5 transcriptions per day, up to 25 MB per file (roughly 25 minutes of MP3 audio). Need more? The Typally desktop app includes generous audio-file transcription plus real-time voice typing in every application, with a 7-day free trial.
MP3, M4A, WAV, WEBM, OGG, OPUS, FLAC and MPEG audio. If your recorder produced it, it almost certainly works.
Up to 25 MB per file — about 25 minutes of typical MP3 or a couple of hours of highly compressed voice audio.
Dozens — English, Spanish, German, French, Portuguese, Lithuanian, Japanese, Chinese and many more, with automatic language detection.
No. Audio is transcribed and discarded immediately. Nothing is saved on our servers.