Say it. See it.
A conformer speech model transcribes your microphone or an audio file in 25 European languages — running entirely in this tab. The recognizer is a 347 MB download from Hugging Face, fetched once, cached forever. Your audio never leaves the device.
Record from your microphone
or drop an audio file on this card
Drag & drop audio here — mp3, wav, ogg, flac, m4a, webm…
The transcript will appear here.