Abstract: With the proposal of emission peak and carbon neutrality, the region integrated energy system (RIES), which can realize the multi-energy interaction and the clean energy consumption, has become a research hotspot. To take into account the interests of different subjects in this system, this paper proposes a two-level game optimal operation strategy of the RIES considering the wind and solar uncertainty and the carbon trading. Firstly, a two-layer multi-agent system model with the energy sales agent (ESA), the energy operation agent (EOA) and the energy use agent (EUA) as the main body is established. Secondly, the Monte Carlo method is used to generate the random scenes and the scene reduction is carried out by the k-means clustering method and the pre-generation elimination technology for the uncertainty of the wind and solar output. Then, considering the operating income and the carbon trading cost of each agent on the source side and the comprehensive demand response on the load side, a two-level game framework of the vertical Stackelberg game nested with the horizontal EOA non-cooperative game bidding is constructed and its Nash equilibrium is proved. Finally, the improved double mutation differential evolution algorithm combined with the YALMIP toolbox and the commercial software CPLEX is used for the simulation analysis in the MATLAB. The results of the example show that the optimal strategy proposed in this paper is effective, which is able to improve the interests of each subject and reduce the rate of the abandoned wind and solar power and the carbon emissions.