A real-time pricing mechanism under the maximization of social welfare does not consider the benefit independence of the power system entities, so it cannot achieve the benefit and utility balance of the power system entities. On the other hand, it may cause the generation or transfer of user load peaks and valleys. In this paper, a two-stage equilibrium model of real-time pricing strategy is established by integrating "load-utility" two-stage equilibrium into the real-time pricing optimization of the smart grid. Based on the revenue utility function of the main body of the power grid system, the pricing mechanism of the purchase electricity price and the real-time electricity price is derived, and on the basis of the pricing mechanism, this paper designs an algorithm based on a multi-objective genetic algorithm (NSGA-II). Finally, a simulation shows that the real-time electricity price pricing mechanism under the two-level equilibrium model can better ensure the fairness of the benefits of the participants in the grid system, and make the user load more balanced. This work is supported by the National Natural Science Foundation of China (No. 72071130).