Speech to text, live
Say it once. Read it back as it happens.
Stream straight from your microphone over a live connection, or drop in a recording — Speech to Text turns audio into a transcript in real time, using Google's streaming recognition and OpenAI's Whisper.
No install. Sign in with a one-time code and start talking.
From a sentence to a transcript
Four steps, and the only one you do is the first.
-
Speak, or upload
Start talking into your microphone, or drop in a .wav recording. Both land on the same screen.
-
Audio goes out over a live connection
Microphone audio streams over WebRTC as you talk; an uploaded file is posted once, as a whole.
-
Google and OpenAI do the listening
Live audio is streamed to Google Cloud Speech-to-Text; uploaded recordings are transcribed by OpenAI's Whisper model.
-
Text lands as it's heard
Interim words firm up into final lines in the transcript, continuously, with no page reload.
Built on the engines doing the actual work
Two ways to get audio in
Talk live over WebRTC, or upload a .wav file. The transcript view is the same either way.
Real engines, not a mockup
Google Cloud Speech-to-Text handles streaming recognition; OpenAI's Whisper handles uploaded recordings.
A proper peer connection
Microphone audio travels over WebRTC, the same transport behind video calls — not a chain of file uploads.
In with a code, not a password
Sign in with a one-time code sent to your email. Nothing to remember, nothing to reset.
Hear it work on your own voice
Sign in, press start, and talk — the transcript fills in as you go.
Open the live demo
