Feature Request: Enable API Access for GPT Builder Models

GPT builders
1

I’m writing to request a feature that I believe would be immensely valuable to developers and users of the ChatGPT platform: API access to Custom GPTs created via the GPT Builder.

The GPT Builder is an incredibly powerful tool that allows developers and non-technical users alike to create tailored GPTs with specific configurations, data storage, and specialized capabilities. However, the lack of API access to these Custom GPTs limits their usability outside the ChatGPT platform’s UI. This creates challenges for developers who want to integrate these customized models seamlessly into their applications.

Use Case

I have created a Custom GPT that performs specialized tasks for my application. While it works perfectly within the ChatGPT platform, I cannot provide access to it via my own app without resorting to less elegant workarounds, such as embedding the ChatGPT UI in an iframe. This limitation prevents me from:

  • Fully integrating my Custom GPT’s functionality into my app’s workflow.
  • Allowing my users to interact with the GPT directly, without requiring them to log into ChatGPT.
  • Automating interactions with the GPT in my app for tasks like pre-processing or post-processing user inputs.

Benefits of API Support

Enabling API access for Custom GPTs would:

  1. Expand Accessibility: Allow developers to embed Custom GPTs directly into their applications without relying on the ChatGPT UI.
  2. Increase Adoption: Encourage broader use of Custom GPTs, especially in enterprise environments where API-driven workflows are a standard.
  3. Simplify Developer Efforts: Reduce the need for complex alternatives like RAG (Retrieval-Augmented Generation) to replicate GPT Builder configurations manually.
  4. Leverage GPT Builder’s Full Potential: Unlock the ability to use the configurations and stored data directly through the API, maintaining the ease of customization provided by the GPT Builder.
  5. Chargeable API calls, simpler access to customised GPT via API embedding

Proposed Solution

Provide an endpoint in the OpenAI API that allows developers to:

  • Specify a Custom GPT ID (e.g., g-xxxxxxxxxxxx) in the model field of an API request.
  • Retain all the behavior and configuration of the Custom GPT created in the GPT Builder.

This would bring parity between the ChatGPT UI and OpenAI’s API, allowing users to seamlessly transition between the two environments.

Conclusion

I strongly believe this feature would significantly enhance the usability and flexibility of the GPT Builder. It aligns perfectly with OpenAI’s mission to make AI accessible and useful to everyone. I would greatly appreciate your consideration of this request and would be happy to provide more details or examples of potential use cases.

Thank you for your time and for continuing to improve this groundbreaking platform!

3 Likes
2

100% NEEDED
I’ve built a customGPT that will be of great value to my students, but not with the current message cap.
I need API access so my students can use my CustomGPT in a far more intense manner.

1 Like