A short-circuit model for silicon carbide (SiC) metal-oxide semiconductor field effect transistors (MOSFETs) using hybrid modeling method based on artificial neural network (ANN) and improved artificial bee colony (ABC) algorithm is proposed in this paper. In order to improve the search ability of the ABC, particle swarm optimization (PSO) is introduced to the scout bees’search strategy. The improved ABC is employed to find suitable initial parameters for ANN model, which can improve the accuracy of modeling results. Based on hybrid modeling method, the normal working model of SiC MOSFETs is established first. The modeling results of I−V characteristics, C−V characteristics and small signal parameters (gm, gd, etc.) are in good agreement with datasheet, which fully demonstrates the validity of the normal working model. Then the short-circuit model of SiC MOSFETs is further obtained based on the relationship between short-circuit current and junction temperature and normal working model. Eventually, the proposed short-circuit model is verified by device- and circuit-level tests. With its precision and simplicity, the proposed short-circuit model can be used to analyze short-circuit faults in SiC MOSFET simulation circuits and provide assistance for the design of protection circuits.