There is variety scheme when a part with multiple features is processed in CNC machines, and hence, different feature sequencing during processing affects not only productivity but also energy consumption. This paper concentrates on the energy-saving strategy by optimizing the feature processing sequence in the part processing stage through reducing the energy consumption of the non-cutting process. The detailed energy model is established considering rapid feed and general feed path in the X, Y, Z+, Z- directions for analyzing the impact of processing feature sorting on reducing the energy consumption of parts processing. The feature sequencing optimization is carried out under the condition of fixed cutting parameters for specific machining features to better reveal the sequence influence on energy consumption and non-cutting time. Meanwhile, the energy consumption of the non-cutting of parts specifically includes the empty pass and an automatic tool change model, while the normal feed and the rapid feed are established in different moving axis, respectively. Based on the developed model, the genetic algorithm is used to solve the optimal processing sequence and the lowest processing energy consumption. Finally, a cutting orthogonal experiment is executed to collect energy consumption data, analyze the data and fit the data to establish a specific energy consumption model for each processing stage. A case study of a part with nine features is used to optimize sequencing, which shows the effectiveness and validity of the proposed method.