Automatically choose the most suitable model for each agent request—boost performance while cutting costs by leveraging the combined strengths of multiple LLM models.