The urban electric vehicle service network is the basic support system to provide energy for electric vehicle (EV). Evaluation of the service capability of urban EV service network is of great significance to the rational construction of EV service network and the large-scale promotion and popularization of EVs. This paper, based on the cloud model comprehensive evaluation method and the analytic hierarchy process, the three-level nine-index charging capacity index system and the three-level six-index urban charging station weight estimation index system are established. The service capability evaluation of the EV service network is evaluated according to the evaluation result of typical charging station and the weight of the typical charging station in the service network. In order to realize the dynamic evaluation of service network service capability, it is necessary to update the evaluation system parameters in real time according to the development of electric vehicle and the change of charging demand. This paper proposes an EV demand growth model based on the improved gray forecasting method and the support vector machine to predict the number of EVs. Energy efficiency maximization principle is used to predict the charge load of the charging stations. According to the charging forecast results of the charging station, the quantitative parameters of the charging piles needed in the charging station in the evaluation system are updated in real time to form the dynamic assessment system of EVs service network’s services capability.