Comment by okeysmokey

Is this using Whisper or something else? It looks like the site performs multiple steps on the audio and did a decent job of guessing who was speaking.