Traditional microgrids have problems such as lack of interaction among users and low utilization rate of renewable energy. Considering the operation mode of photovoltaic (PV) output and energy storage (ES) in smart buildings under different climatic conditions, this paper proposes a micro?grid operation mode based on multi?agent interaction to optimize the user's energy cost of the micro?grid and improve the ability to absorb renewable energy simultaneously. First, this paper compares and analyzes the classic operation mode of microgrid and the multi?agent interactive operation mode, and studies the energy consumption behavior of each agent in different modes; Then, the contract price is set flexibly to improve the efficiency of the energy interaction between the smart building operators and the electric vehicle (EV) charging station (CS) operators in the microgrid, and the energy flow inside the microgrid under the multi?agant interaction mode is analyzed; Finally, considering the impact of environments on PV output, the energy cost of users in the microgrid under different operating modes and the microgrid’s capacity of renewable energy consumption are analyzed. The simulation results show that the multi?agent interaction mode of the microgrid increases the consumption rate of renewable energy from 67.83% of the classic operation mode to 100% when the light conditions are sufficient, and at the same time optimizes the energy cost of smart building operators and CS operators.