An intelligent API router that automatically selects the best LLM model for each request based on complexity, cost, and quality requirements. Simple requests go to cheap models, complex ones to powerful models. One API endpoint, optimal model selection.