This paper presents a vehicle routing model with multiple warehouses and product arrangements in vehicles. Problem modeling consists of two general stages that the result of the first stage directly impacts the decisions of the second stage. Therefore, in the first stage of the problem, the main goal is the optimal arrangement of products requested by customers in packages. While in the second stage, the issue deals with the arrangement of packages (optimal dimensions obtained from the first stage) in vehicles and the optimal routing of vehicles. The goal at this stage is to reduce transportation costs, locate facilities, and select suitable vehicles. Genetic algorithms have been used to optimize the value of the first stage objective function to solve the proposed two-stage model. The weed optimization algorithm has been used to achieve the optimal value of the aim of the second stage function. The result of small sample size problem calculations is the relative difference of 0.68% of the genetic algorithm with GAMS in the first stage and the relative difference of 0.16% of the weed optimization algorithm with GAMS in the second stage; Which shows the high efficiency of the mentioned algorithms in solving the sample problem. Then 10 sample problems were designed and solved in larger sizes. The results show the high efficiency of genetic algorithms and weed optimization in solving large sample problems and case studies in a much shorter time.