The reliability of mechanical transmission systems and the stability of motion accuracy have a significant impact on the performance of CNC equipment. Existing studies have rarely established accurate reliability optimization allocation models. Therefore, a reliability allocation method that incorporates motion accuracy stability and a generalized reliability-cost function (G-RCF) is proposed. Firstly, the mechanical transmission system is decomposed by using the meta-action theory to obtain the meta-action units (MAUs). The motion accuracy stability, structural complexity, and comprehensive maintenance cost factors of MAUs are analyzed, and the traditional reliability-cost function is modified to obtain the generalized reliability-cost function. Then, taking the minimum generalized cost as target, a mathematical model for optimizing the allocation of system reliability is established. Using the intelligent algorithm to obtained the reliability design values of each MAU. Finally, the effectiveness of the proposed method is demonstrated through an engineering example.