Particle Swarm is a heuristic technique based on collective behavior of birds. Several researches depicts that the PSO suffers from untimely convergence. To defeat the issue of untimely convergence in PSO several solutions are proposed to increase the performance in term of accuracy. This paper suggests a new hybrid mutation operator which used Chi-square and stable distribution. The hybrid mutation operator leads the swarm from local minima to global minima for better solution. To validate the new hybrid scheme, a 12 benchmark optimization functions are used in experiment and compared the result with pervious 6 variants of PSO, proposed variant achieved better results than previous 6 variants.