Abstract: The virtual power plant (VPP) generally reduces the impact of uncertainties, but due to the conservative nature of its scheduling schemes, it is difficult for a VPP to obtain the highest economic benefits without considering the internal randomness in the process of participating in the electricity and gas market. In order to fully exploit the economic benefits of VPPs, a stochastic optimization scheduling model is proposed for electric-thermal-gas VPPs with consideration of the uncertainty of electricity price and wind-photovoltaic under gas and electricity market. The objective function of the model is the total benefit of VPPs, which is the difference between the sales of electricity, heat and gas and the cost for electricity to gas conversion, carbon capture, carbon emission and fuel. The superquantile method is introduced to convert the total benefit optimal model of VPPs with multiple random variables into a super-quantile random optimization model. For the convenience of calculation, the model is further processed into a discretization calculation model, and is solved with the spatial particle swarm optimization algorithm. The simulation results show that the VPP obtains the optimal benefits through optimizing the sale schemes of electricity and gas, and with consideration of various random variables in the process of participating in gas and electricity market, the VPP has more opportunities to obtain higher economic benefits after avoiding risks.