Metal additive manufacturing has been successfully applied in aviation, aerospace, mold, medical and other fields due to its characteristics such as rapid, mold-free, and free-form complex structure. As the complexity and volume of manufactured parts continue to increase, the amount of data in 3D models is increasing, and the time required for data processing has increased significantly, especially the steep increase in path planning, which has become the main bottleneck restricting the application of this technology. This problem needs to be solved urgently. In order to reduce the time required for path planning, based on the basic fact that the contours between two consecutive layers of additive manufacturing are similar, here the two-dimensional contours obtained by the slices of the three-dimensional model are divided into groups according to characteristics, and an innovative inter-layer information inheritance algorithm is proposed. Using the calculated filling path information of the previous layer, the filling path of the current layer can be quickly calculated. This algorithm does not need to calculate the intersections of each scan line and many contour loops of the current layer, which greatly reduces the calculation amount of path filling and speeds up the filling. The experimental results have proved that the overall calculation efficiency of the proposed algorithm is significantly higher than that of the traditional path filling algorithm, especially for the models with a constant cross-section or continuous cross-section, and the acceleration effect of the proposed algorithm is more excellent.