Grid is a computing extensive computational environment that makes the enormous processing power for distributed processing. Grid resource management is important and for agent-based approach is a way to manage its resources. Load balancing is an important section resource management system. Paper model of multilevel load balancing in grid and ARMS environments provided, resource management model provided that the estimate is load node. In the First level of model, using genetic algorithm is used for load balancing nodes between internal resources. That is, in the second level, Add method of the neighborhood and overloaded nodes are added to the rainfall distribution in the limited area around the player. The third level is the ecosystem is one of the intelligent ants, each method is a method of load balancing. Picking methods In a multilevel structure is called MLBM that liminates the disadvantages and satisfies the majority of the load balancing properties. The results show that the model is superior to its predecessors gain in increasing productivity and reduction srbarartbaty.