The increasing number of vehicular networking devices and application demands has made the limited computing and communication resources a significant challenge. The heuristic task offloading strategy mechanism was proposed to improve the efficiency of task offloading in vehicular networks. This strategy mechanism utilized the Nondominated Sorting Genetic Fireworks Algorithm (NSGFA) based on the characteristics of the problem, integrated consideration of the multi-target balance between system offloading costs and load balancing. Experimental data results demonstrate that this method performs well in reducing latency, energy consumption, and load balancing, effectively enhancing the quality of service for vehicular network users.