llm-switchboard

Blazing-fast, zero-cost local LLM router. Classify and route prompts to specialized AI models (OpenAI, Claude, Gemini, Llama) with <1ms latency using heuristic rules.