Ask HN: What changes did you see after shifting to GPT-4o Model in your wrapper?

Seems more responsive than 3.5 did with higher quality. Excited to learn more and be able to integrate the real-time voice features.