Longer output in front-end or return via back-end/API?

Hi all,

I just developed a GPT that allows corporate professionals to analyze Corporate Policies and it develops Summaries, FAQs, Definition lists, etc.

You can find the GPT via my site https://martinsgpt.com

However, I have the problem that it can only process a little bit of output and then the user each time needs to ask to continue. Probably because of the token limit. Can I reconfigure the GPT so it does a few cycles in the background (eg up to a certain limit, eg. it developed 100 questions) and then outputs in the front-end or maybe even have it send a completed document via an API to an email address ? I was thinking of having it send to the Assistant in the background but that means users will use my account ($) instead of their own.

All suggestions welcome.

Best regards,
Martin