Agent: Cursor, Claude CodeLLM: Claude 3.5, GPT-4#local-llm#model-proxy#llama-cpp#ai-infrastructure#developer-tools
A lightweight Go-based proxy that lets you run multiple local LLMs simultaneously and switch between them on demand. Works with any OpenAI/Anthropic-compatible server (llama.cpp, vllm, etc.), making it perfect for developers experimenting with different models in their AI workflows.