April 29, 2026 – LifePrompt, a well – known Japanese AI company, has recently made public some remarkable test results regarding ChatGPT, which is based on OpenAI’s latest model.
ChatGPT has demonstrated outstanding performance in the undergraduate entrance exams of both the University of Tokyo and Kyoto University, surpassing the scores of top – scoring human candidates in multiple subjects.
In the highly competitive entrance exam for the medical program in the natural science category III at the University of Tokyo, ChatGPT outperformed the human top – scorer by a significant margin of 50 points. What’s more, it achieved a perfect score in the mathematics subject. This is a sharp contrast to 2024 when the AI failed to pass all the entrance exams of this university.
In addition to the exams at the University of Tokyo, ChatGPT also took part in this year’s national unified college entrance exams in Japan. LifePrompt has compiled and analyzed its overall scores. For the humanities and social sciences exam at the University of Tokyo, with a full score of 550, ChatGPT scored 452. In the science exam, also with a full score of 550, it obtained 503 points. Both scores are higher than the highest scores of the admitted students announced by the university. The highest score in the humanities and social sciences category III was 434, and in the science category III, it was 453.
At Kyoto University, ChatGPT also showed its strength. In the law school entrance exam, it scored 771 points, exceeding the highest admitted score of 734. In the medical school exam, it got 1176 points, higher than the 1098 points of the human top – scorer. These results indicate a significant leap forward in the AI’s ability to handle complex academic exams.
