To get a natural response from GPT i am sending the JSON response to GPT -4 and getting thr response in a natural language format. It works good but has a very high response time

To get a natural response from GPT i am sending the JSON response to GPT -4 and getting the response in a natural language format readable by user. It works good but has a very high response time.Can you suggest me some other ideas where i can achieve the same but in lesser time. it takes 18seconds and sometime even 28 seconds to give the response

Hi Suhail and welcome to the forum

Please provide the full context: prompt, model, parameters