also LLM providers optimize your request routing to a cheaper model sometimes, I wonder if there is a way to structure prompt that will route to a large (more e

also LLM providers optimize your request routing to a cheaper model sometimes, I wonder if there is a way to structure prompt that will route to a large (more expensive, and arguably better) model for better results