Demand response (DR) is important to account for behaviors of the demand side to yield an optimal dispatch result. However, it is difficult for energy suppliers to collect customers’ private information unless there is an incentive mechanism for customers to do so. Therefore, this paper proposes a new integrated generation–consumption dispatch based on compensation mechanism considering DR behavior. Firstly, in light of the dayahead load forecast data, we deduce the utility function model of different customers. By subtracting generating units’ operation cost from consumers’ total utility, the dispatch model have a decentralized demand participant structure based on this utility function. The utility function is used to describe consumers’ preferences and energy consumption behaviors. Secondly, an effective compensation mechanism is designed to ensure customers to select the level of compensation appropriate to their willingness to curtail load. Finally, a new dispatch model is proposed that incorporates the DR compensation mechanism into the generation–consumption dispatch model. The new model can improve the interaction of generation and consumption, and benefit both the energy supplier and its customers. The proposed model is piecewise linearized and solved by a mixed-integer linear programming method. It is tested on a six-bus system and the IEEE 118-bus system. Simulation results show that the proposed model can realize both maximum social welfare and Pareto optimal results.