Yes, I tried it. It's good enough that you can tell me Similarly, can you suggest a good model for text-to-speech and speech-to-text that looks a realistic?