Whisper Transcription Demo

Run Whisper transcription on audio or YouTube video. Whisper is a general-purpose speech recognition model, trained on a large dataset

Model size(s)

turbo

Return word timestamps

Decoding temperature

0 1

Average log-probability threshold

-10 0

No-speech probability threshold

0 1

Compression ratio threshold

1 5

Upload or record audio

... or paste a YouTube URL (audio only)

Transcription Results

Transcription Results