Hi there! I’m discussing the client-side of the GPT-4 chat. I find that the API still tends to be rather slow; I’ve not noticed a significant improvement in this area. Additionally, the costs remain rather high… This prevents its widespread use in applications, unfortunately