In the process of hobbing motion chamfering of cylindrical gear tooth profiles, the tool parameters and mounting position parameters of the chamfering tool directly affect the final chamfering result. Aiming at the multi-objective optimization requirements of tool parameters and mounting position parameters of the chamfering tools, a co-optimization method of multi-parameters based on Improved Non-dominated Sorting Genetic Algorithms-II (INSGA-II) is proposed. Firstly, considering the consistency and symmetry of chamfering result, the co-optimization model of multi-parameters is established with the minimum variance of chamfering width and thickness as the objectives, and the tool parameters and mounting position parameters as variables. Secondly, to avoid the tool parameters and mounting position parameters falling into local optimum, the normal distribution crossover operator is introduced in Non-dominated Sorting Genetic Algorithms-II (NSGA-II) to improve the searchability, and INSGA-II and Fuzzy C-means_Grey Relational Projection (FCM_GRP) are used to optimize and decision parameter combination. Finally, to further verify the applicability of the proposed co-optimization method, the optimized parameter combination is used to calculate the tool rake face profile, and the local curve superposition is used to improve the uneven chamfering at the transition arc. Based on the simulation of tooth profile chamfering, the simulation results verify the effectiveness of the proposed method.