Model charges in the Assistant

I am interested in using an Assistant in production with the pinecone-assistant repo.

  1. How do you charge for the cost of using the gpt-4o or claude-3-5-sonnet models in the Assistant?
  2. Is there a way for better control of the model and search parameters, and the chat history when using an Assistant from an application?

I am also interested how to choose a model in API when creating an assistant. It seems that there is no such information in the documentation