Free tools Windows power users keep installed
One-click scans. No signup required.
Yes—but only in a narrow sense. In 2017, a system developed by Tsinghua University and iFlytek scored 456 on the written portion of China’s clinical physician qualification examination, above that test’s 360-point pass threshold. It passed a written test; it did not receive a medical license or demonstrate that it could practice medicine independently.
What exam did the robot take?
The system, called “智医助理” (Medical Assistant), was jointly developed by Tsinghua University’s Department of Electronic Engineering and iFlytek. The team set participation in the 2017 clinical physician qualification written examination as a development milestone, according to Tsinghua University’s November 9, 2017 account.
As an Amazon Associate I earn from qualifying purchases.
This was a robot-specific test of the clinical-category written paper, not a claim that an AI candidate completed every requirement for physician qualification. The exam’s format and content figures below describe the 2017 test only; they should not be taken as specifications for China’s current examination.
What score did it get?
Tsinghua reported a score of 456 points, against a 360-point pass threshold for the written test. Xinhua’s November 7, 2017 report also reported that the system passed with a score of 456.
#1 Best Overall
Tsinghua described that year’s written paper as 600 objective questions across four timed units, covering basic medicine, medical humanities, clinical medicine and preventive medicine, with about 2,700 knowledge points. The university said about 530,000 candidates took the national examination that year. Those numbers provide context for the reported result, but the score alone does not show how the system performed in a clinical setting.
How was the test conducted?
The system answered the clinical written paper on August 26–27, 2017, during the examination period. Tsinghua said it sat in a designated room under the observation of exam-center invigilators and notaries; the process was recorded, and the room had no network or signal. Xinhua likewise reported that the system answered the same paper as human candidates, at the same time, in a room without internet access or signal.
That supervised setup supports a specific conclusion: the system produced answers on the written examination under controlled test conditions. It does not establish that the machine completed the separate practical-skills assessment or that it could examine, communicate with, diagnose or treat patients.
Did passing mean the robot was licensed to practice?
No. The reported 456-point result was for the written comprehensive test. It did not, by itself, confer a physician qualification certificate, establish a pass on the separate practical-skills assessment, or authorize the system to provide medical care. A written knowledge test and the broader demands of clinical practice—including practical skill and patient interaction—are different measures.
A 2018 Xinhua report described a related iFlytek AI general-practice assistant beginning a trial at a community hospital. It said the system could help review medical histories and suggest prescriptions, while doctors still had to sign diagnostic reports and prescriptions. That report documents a trial at that time, not the system’s present availability or deployment.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.What technology did the developers say they used?
Tsinghua’s institutional account attributes the system’s exam performance to methods including semantic tensors and multi-semantic deep learning. It also describes reasoning based on key-point semantics, context and evidence chains. These are the developers’ description of their approach, not an independent evaluation of the system’s technical design or clinical effectiveness.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
