Text this: Comparative Performance of Large Language Models on European Gastroenterology Board-Style Questions