Optimum selection of cutting conditions importantly contribute to the increase of productivity, quality and the reduction of costs. Currently, in countries around the world and has a lot of research to select the optimal cutting conditions on CNC machines. However, these studies often use the popular methods such as differential method, regression analysis, linear programming method, the method of surface targets surface. This .paper gives a research using artificial neural network (artificial neural networks) for multi-objective optimization. which is the surface quality (Ra), production costs (Cp) or processing time (Tp). Combining empirical research to select the optimal cutting when machining hardned 9XC alloy steel with PCBN indexable inserts.