curl -s http://localhost:8000/v1/chat/completions -H 'Content-Type: application/json' -d '{"model":"meta-llama/Llama-3.1-8B-Instruct","messages":[{"role":"user","content":"Hello"}]}'

Calls vLLM's chat endpoint with one user message, exactly like OpenAI's API, and prints the assistant's reply as JSON. The messages array carries the whole conversation history on multi-turn calls. Add "stream":true to receive tokens incrementally instead of one big response.