Selective laser melting (SLM) is a prevalent additive manufacturing (AM) technique for the fabrication of metallic components. A modified GTN (Gurson-Tvergaard-Needleman) model was developed, based on the understanding of the SLM process and SLM-manufactured parts, in order to characterize void growth and void shear mechanism to predict the ductile fracture behavior of SLM-fabricated Ti6Al4V alloys under uniaxial stress states. The effect of the number of hidden layers and neurons, as a basic parameter of an artificial neural network (ANN), on predicting parameter relation accuracy was investigated. In this study resulted due to the complex relation among GTN fracture parameters and fracture displacement, defining more hidden layers in ANN improves the accuracy of predicting the damage and fracture behavior of SLM-fabricated Ti6Al4V alloys under uniaxial stress states; however, forecasting maximum force is achieved accurately by fewer hidden layers in comparison with fracture displacement needing to higher layers to predict precisely. Furthermore, the system R 2 -value reaches higher accuracy more than 0.99 for both maximum force and fracture displacement based on selected hidden layers and neurons.