Assessment of the Ability of the ChatGPT-5 Model to Pass the Endocrinology Specialization Exam

    September 2025 in “ Cureus ”
    Ada Łątkowska, Piotr Sawina, Tomasz Dolata, Dawid Boczkowski, Aleksandra Wielochowska, Anna Kowalczyk, Michalina Loson-Kawalec, Dominika Radej, Wojciech Jaworski, Weronika Majchrowicz, Melania Olender, Jaroslaw Adamiak, Julia Sroczynska, Riham Suleiman, Julia Glinska, Patrycja Szczerbanowicz, Patrycja Dadynska
    TLDR ChatGPT-5 can pass the endocrinology exam but isn't fully reliable for clinical decisions yet.
    The study evaluated the ChatGPT-5 model's performance on the Polish endocrinology specialty exam, consisting of 120 multiple-choice questions. ChatGPT-5 achieved a score of 76.47%, correctly answering 91 questions and surpassing the 60% pass threshold. This suggests that large language models like ChatGPT-5 could support postgraduate medical education and exam preparation. However, its reliability in clinical decision-making and handling ambiguous cases remains uncertain, indicating a need for further research to explore its potential role in clinical practice.
    Discuss this study in the Community →