The inventory routing problem simultaneously considers both the inventory problem and the delivery problem; it determines the amount of delivery of inventory and the delivery route such that the total cost is minimized. In this study, we use stochastic programming to consider a multi-product inventory routing problem that considers the demand variation throughout multiple periods. Determining the delivery route for each period is difficult. Therefore, we propose a model that fixes the delivery route throughout the planning period and compare the calculation results to prove its practicality. In addition, the problem in this study is an integer programming problem, and solving a large-scale problem using the direct method would be time-consuming. Therefore, we apply the accelerated Benders decomposition method, which combines two cuts: the optimality cut with the solution of the linear relaxation problem, and the Pareto-optimal cut. We demonstrate its effectiveness through numerical experiments.