cgaeking

Smart LLM router — save 78% on inference costs. Routes every request to the cheapest capable model across 30+ models from OpenAI, Anthropic, Google, DeepSeek, and xAI.