Generating text using the 'n' parameter

Actually not like that, lets say we want to generate emails for 10 different users, so the prompt context will be different for each but company context would be same, so instead of making 10 different OpenAI calls, we want to reduce it to least possible?

Interesting. Even in that scenario, each of your final prompts are different right because the content changes each time.

If you have sufficient tokens, maybe you could alter the prompt to stuff context for two users in a single prompt.

Or even more, and have the model return the response in a JSON format or something