Abstract: This paper investigates the nesting optimization problem within the glass cutting process of a large-scale glass manufacturing enterprise. The objective is to address issues such as the lack of automated nesting capabilities in the existing scheduling system, the inefficiency and poor quality of manual nesting, and the need to improve material utilization rates. To tackle the optimization challenges arising from complex and varied glass dimensions, a grouped genetic search algorithm incorporating constraint programming is proposed. Addressing the limitations of standard placement algorithms—such as the Lowest-Level-Line algorithm and the Remaining Rectangular Space method—this approach utilizes constraint programming to generate diverse initial solutions that serve as the starting population for the search algorithm. Additionally, a classification strategy distinguishing between “superior” and “ordinary” populations is implemented, enabling the optimization of the ordinary population while maintaining the quality of the superior group. By combining the rapid feasible-solution generation capabilities of constraint programming with the global optimization strengths of intelligent algorithms, this method effectively resolves the complexities caused by the fragmented and diverse glass sizes found in orders. Finally, the algorithm was integrated into the enterprise's scheduling system and tested using real-world order data. The results demonstrate an increase in average material utilization from 75% to 86%, with a peak utilization of 89%, thereby validating the method's effectiveness in actual production scenarios.