# llm-router

routes between self hosted LLM servers. All servers should be exposing the same model and running on the same server software. I test this with lemonade and qwen3.8 specifically