Gradient-free metaheuristic algorithms have demonstrated excellent performance in solving complex optimization problems and have attracted extensive research attention. The nutcracker optimization algorithm is a metaheuristic algorithm proposed in recent years, which is highly regarded for its excellent performance. However, it still has some room for improvement in solving high-dimensional optimization problems. Multi-population strategies are a widely used technique for enhancing the performance of metaheuristic algorithms. This approach is readily integrated with a multitude of swarm intelligence and evolutionary algorithms, and it serves to forestall the occurrence of premature convergence phenomena. In light of these findings, this paper puts forth a novel multi-population nutcracker optimization algorithm, which is evaluated on the CEC 2017 benchmark function set and compared with other algorithms. The experimental results demonstrate that the incorporation of multi-population enhances the convergence speed and accuracy of the nutcracker optimization algorithm.