Pick an audio or video file and hit Transcribe. It's sent to a fast, GPU-accelerated transcription pipeline, then a plain-text transcript is ready to preview and download. Supported formats: MP3, FLAC, WAV, M4A (audio) and MP4, MKV, WebM, MOV (video).

Stage 1 (current) does not label who's speaking - it's one continuous transcript. Speaker labelling (diarization) is planned as a later option for recordings with multiple people.

Every transcript you make is saved here, not just the one you're currently viewing - come back any time to re-download a transcript, grab the original file, rename an entry, or delete it.