As digital transformation penetrates into residents' daily lives, online services have changed residents' living habits and community services, which has a remarkable impact on the efficiency and fairness of social service configuration. Current research has been limited to traditional offline scenarios of community retailing and the impact mechanisms, and less consideration has been given to the impact of the growing online to offline (O2O) platforms on the accessibility and fairness of community retailing. In addition, the methods for assessing the accessibility of community retailing under the online scenario should be optimized. Taking Nanjing City's community retailing as an example, this study integrated Web Map Service (WMS) and Scenarios Accessibility Gap (SAG) to map the spatial pattern of accessibility of community retails under offline, online, and integrated scenarios based on offline and online retail characteristics. The Spatial Lag Model (SLM) and Gini coefficient were adopted to measure the fairness of community retail accessibility for different income groups under the three scenarios. The results show that: 1) Under all scenarios, community retail accessibility in Nanjing shows an urban-rural polarization pattern, decreasing gradually from the central city to the other parts of the city and then to the suburbs. 2) Of all communities, 96.76% have improved accessibility of retail services by online services, and 64.23% of the blind spot communities under the offline scenario have improved assessibility with online services that significantly optimized the accessibility of traditional offline community retail. 3) Under the offline scenario, retail accessibility follows an inverted U-shaped curve as the housing price rises; under the online scenario, retail accessibility increases linearly as the housing price increases, which indicates that online services changed the relationship between economic attributes and retail accessibility in offline communities, amplifying the inequality of opportunities between communities with different economic attributes. 4) Under the offline scenario, the gap between community retail resources of different income groups is relatively small (0.20), while the gap is larger and inequity is higher under the online scenario (0.24); due to the scale effect, the fairness of community retail within most of urban district is higher under the online scenario compared to the offline scenario. The findings can provide some references for the construction of community living circle and ensuring the fairness of community services.