Aiming at the task in which multiple loitering munitions should penetrate the defense and attack the target under time constraints, a cooperative path planning algorithm is designed based on Multi-Hierarchy Grey Wolf Optimization. Firstly, the model of the cooperative path planning problem is built, and cooperative cost function is set. Secondly, a multi-hierarchy structure is constructed, which provides various evolutionary directions to grey wolves’ population, including cooperativity, optimality and randomness. Finally, Taboo Search and dispersal mechanism are added into the algorithm to expand searching area and elevate optimization ability. Simulation work for attacking task from all directions has been done, and comparison between Multi-Hierarchy Grey Wolf Optimization and other general algorithms has verified its advantage in rapidity, optimizing and completeness, and shown that its specific architecture gains better fitness for cooperative path planning problems.