Copilot runs rival AI models on the same question to spot agreements and differences

Copilot's Model Council sends one prompt to Anthropic's Claude and OpenAI's GPT models simultaneously. A judge model summarizes agreements, divergences, and key insights. This eliminates manual copying between apps, saving time on comparisons for research or analysis where model reliability matters.