Is it possible to access a ChatGPT GUI-created custom GPT through the API?

Below is some guidance on this topic:

Custom GPTs in ChatGPT GUI vs. API Access

Currently, the customizations you make on the ChatGPT GUI (often referred to as "custom GPTs") are not directly exposed via a dedicated API endpoint. The API lets you interact with OpenAI’s chat models (like GPT-4) via the chat completions endpoint, where you can supply your own system and user messages. In essence, you can replicate many aspects of your custom GPT by crafting similar instructions and prompts within these API calls.

How to Get More Information

• API Documentation: For details on how to construct a request, including setting system instructions, check the chat completions documentation page on OpenAI’s official API reference. This documentation provides code examples (cURL, Python, Node.js, etc.) showing how to format your request and handle the response.

• Release Notes: For updates regarding customizable GPTs and related features, you can refer to the GPT release notes section on OpenAI's documentation site.

Summary

While you cannot directly access the exact custom GPT created via the ChatGPT GUI using the API, you can use the chat completions API to simulate a similar behavior by providing context, instructions, and user messages that mirror your customizations. Always refer to the official API reference for the most up-to-date information on how to format your requests.