How to update models on LLM agents?

I got tired of explaining this over and over both to colleagues as well as users, so here goes a fix