I was on the Praat homepage and I noticed that Anastasia Shchupak was added as an author to Praat. After a bit of digging, I saw on the Acknowledgements page that she is responsiblle for automatic speech recognition with word-level timestamps, automatic speaker diarization, and adaptation of whisper.cpp, Silero VAD and pyannote.audio. Whoa. I hadn’t heard that Praat can now do automatic transcription. So, I thought I’d put together a quick tutorial to see how to do that and to see how it works.
Looks like you need to install Whisper. See documentation on how to do that here.
https://www.fon.hum.uva.nl/praat/manual/TextGrid_Sound_Transcribe_interval.html
Find preferences folder here. I had to manually create it.