Allows for self hosting & utilizing inferencing services for models like LLama, Deepseek, Grok, Claude, etc. Billing should be adjusted on million / tokens to user defined allowing you to charge the rate your paying or even up-charge.

EXTREMELY EASY IMPLEMENTATION:

Just allow the user to set the OpenAI Base URL to the endpoint location. Here is a doc showing you how to use Together.ai's endpoint: https://docs.together.ai/docs/openai-api-compatibility

PS: Open Source your code base and I would have done this myself. I am available to help implement this as we have our AI agents we want to implement inside of GHL.