Doesn't seem like there's a demo set up by them yet, but you can just download the model weights and run the inference yourself and compare it to OpenAI Whisper or anything you have access to.

yeah thanks. just saw that. fingers crossed it's really good for korean. i'm just glad they released it at least. the other sota TTS models by big companies won't be seeing the light of day.

reproducing the english TTS demo resulted in very underwhelming quality.