As dynamic pricing is one of the most recent topics in the energy management system in smart grids, prosumers are invited to participate in the trading mechanism. Unlike the current grid, the bidirectional communication enables the information and energy transfer between the entities. Therefore, the complex dilemma, between selling energy and consuming it, outcomes specifically from the prosumer side. This paper proposes a hierarchical solution based on game theory to identify the optimal local consumption and surplus selling of energy while maintaining an agreement between the sellers, the buyers, and the grid. Our method is composed of three mechanisms: (1) the interactions between the sellers and the grid to define the optimal energy consumption, (2) the interactions between the sellers and the buyers to identify the price that maximizes the prosumers gain, (3) the interactions between the grid and the buyers to afford the rest of the needed energy when the second step is not sufficient. The proposed method based on Stackelberg game, takes into consideration the variety of the prosumers present in the energy market system. Depending on the priority accorded to each one and the spatial constraints, an optimal energy demand is defined using Lagrange multipliers. A distributed algorithm is then proposed to manage the complex interactions while preserving the minimum exchange of information, guarantying the privacy of the participants. Extensive simulations are carried out to verify the performance of the proposed method while maintaining a fair energy distribution. Thus, a unique equilibrium for the proposed game is proved.