Consult multiple external models on the same task and compare their answers. Use when the user wants outside opinions, when a decision benefits from model diversity, or when one model's answer should be checked against a small panel before acting.
