Is it possible to create dedicated APIs for specific "Personal" GPTs?

Hey guys,
Fist of all, I wanna appreciate your great work on the new “Personal GPTs” projects, honestly, didn’t see this feature coming this soon.
also thanks for the quiet close release dates for GPT 4 128K.

you have done this amount of great job in such a short time, and therefore, I had an idea which could help me and many others like me. is it possible for you guys to create a platform to generate dedicated APIs for each personal GPT ? I mean, we need to be able to call a specific GPT that we trained in our codes though some kind of API that is only generated for that specific GPT.

that would be great to hear back from you guys
Best Regards
Pooya “Yooz” Ashtari

GPTs are a function only available in ChatGPT Plus.

They act as a replacement to plugins, or rather an enhanced “chat share” which has been disabled from being continued to reduce utility and overlap of sharing conversations.

They are not an API function. They are an ad to upgrade to ChatGPT plus when you share them.


API programmers can program with AI models to do all similar inference tasks using the models directly - and can integrate into their own paid services, paying for the usage of the language data.

thanks for your reply,
I know these two function seperately, but my question was, isnt it possible to integrate GPT+ with the function of API generating platform, and it would use the extra credit, paid on the OpenAI account ?

No, about the only thing that might transfer is your learning of what instructions might make the AI act in a particular way.

There are two separate billing systems and there is no “extra credit” from having a ChatGPT Plus subscription.

An API developer does not need ChatGPT Plus to use the API, and would quickly find that there is little use for ChatGPT and the Plus system of enticements when they can make their own AI-based products. ChatGPT is the competitor, relevant only when OpenAI makes features such as DALL-E-3 exclusive or prohibitively expensive otherwise.

You’d also find that the premature assistants platform - if you’re the one paying for all the data it can consume without controls - is not the best way to implement an AI solution.