Purpose/Significance To study the typical characteristics of urban residents when they choose internet medical services, and to establish patient behavior portraits, so as to provide decision-making basis for improving internet medical services. Method/Process Taking patients in Beijing as the research object, patient group portraits are constructed from three dimensions:individual attribute characteristics, online consultation behavior and medical demand preference, and online patient groups are classified by the clustering method. Result/Conclusion Online patients are classified into five distinct groups:value potential, platform loyalty, potential conversion, strong demand, and service transfer. Overall, online patients from major cities tend to have lower case severity and lower levels of activity. They typically prefer lower-cost text-based consultations and show a preference for doctors from top-tier hospitals with higher professional titles and superior online service quality. These five different patient groups possess their own typical characteristics, and specific measures need to be taken to meet their personalized medical needs.