Hello! I've got a working server endpoint that correctly receives messages from vapi, passes them to my custom llm, and generates a response, however I don't know what format to send the response back to vapi. Can I get some help with this?

yes check example impl here: https://github.com/VapiAI/server-side-example-serverless-vercel/blob/master/api/custom-llm/openai-sse.ts