Is it possible to access a custom GPT from the API?

Is it possible to access and integrate a custom GPT model, specifically tailored for individual requirements, through the OpenAI API?

1 Like

Hi Shane

Could you elaborate what exactly you mean?

By “custom GPT model” do you mean a GPT (as of GPTs Introducing GPTs) or you mean a fine-tuned GPT model?

In the Chat GPT I see this:

I want to send calls to these custom GPTs…I don’t think there is a way to do this but it would be cool if there were.
Thanks,
-Shane

Short and long: No.

GPTs are for use within ChatGPT Plus, exclusive use to subscribers of the service. They also may rely on other developer’s APIs, authentication, and subscription.

You would need to develop your own code-based solution using API models that offer the same techniques – and pay the bill for all language data yourself – to have an application you can integrate in your own software or site.

2 Likes

No, that’s just a service of ChatGPT.

Screenshot from 2023-12-06 13-05-40

But there’s a service that can do exactly the same thing.

Create an assistant. It’s also possible to attach files and give instructions.
You can import the id of the assistant you created and use it.

You can create a GPT using Assistant api with a project like this and create an api

2 posts were split to a new topic: Trouble Converting Custom GPT to Assistant