As a wireless power transfer (WPT) system operates for a long period, especially with the presence of coil misalignment, the overheating of the magnetic coupler could reduce stability and rise safety risk of the system. To this end, this paper proposes a digital twin model for the magnetic coupler, with the objective of predicting the coupler’s thermal distribution under various operating conditions. The digital twin model is able to predict thermal change of the magnetic coupler, thus the WPT system can lower or turn-off the power transfer in advance to avoid overheating. Also, it provides a valuable and convenient tool for the optimization of the magnetic coupler considering factors like coil misalignment, ambient temperature fluctuation, and load variation.