Will code one day run a code? Performance of language models on ACEM primary examinations and implications.

Large language models, by passing the ACEM primary examination, show potential as tools for medical education and practice. However, limitations exist and are discussed.