AI Agent Model Input Suggestions

The concept is:

Making an HTTP request to integrate with a custom LLM model as input for an AI agent.

My use case:

I have been utilizing a third-party LLM that is not currently listed in the LLM mode options and cannot be run locally (meaning the AI starter kit cannot be used). I am accessing this LLM via an HTTP request. It would be advantageous if the AI agent could accept HTTP requests to these LLM models as input.

I believe adding this would be beneficial because:

It significantly expands the possibilities for testing with various LLMs.

Any resources to support this?

For example, Perplexity AI, and many others.

Are you willing to work on this?

Yes