Abstract: To address the power measurement challenges of perovskite photovoltaic (PV) modules arising from their inherent metastability, this study proposes a rapid prediction method based on kinetic modeling. Current testing protocols require extended light soaking to achieve steady-state conditions, severely limiting high-throughput characterization efficiency in both laboratory and industrial settings. Accordingly, three commercial perovskite PV modules exhibiting distinct power stabilization behaviors were selected to evaluate the applicability of logarithmic and bi-exponential models in characterizing power dynamics. The results demonstrate that the bi-exponential model, by incorporating two independent time constants, effectively decouples and captures the complex kinetic processes driven by ion migration and carrier recombination. Critically, this model enables precise extrapolation of the final Stabilized Power Output (SPO) using only approximately 1.5 h of light soaking data, with the relative prediction error maintained below 1%. This work validates the feasibility of replacing full-duration light soaking with kinetic modeling, providing a robust theoretical foundation for establishing efficient stability assessment benchmarks for perovskite modules.