Reliable model swapping for any local OpenAI compatible server - llama.cpp, vllm, etc - View it on GitHub