<p>Regions with strong earthquakes and heavy rainfall are frequently affected by the coupling of seismic and rainfall effects, leading to frequent and highly destructive landslides that severely threaten regional safety. There is an urgent need for accurate landslide susceptibility assessment methods to support disaster prevention and mitigation efforts. Existing data-driven methods are limited by insufficient understanding of landslide mechanisms and reliance on historical data, while physical analysis methods mostly focus on single triggering conditions, making them unsuitable for complex coupled environments. To address these issues, this study proposes a Rainfall-Earthquake Model (REM), which integrates the seismic permanent displacement model and steady-state hydrological model, and incorporates vegetation's soil-reinforcing effect to enhance accuracy. The model does not require training with historical landslide data and enables rapid Landslide Susceptibility Prediction (LSP) immediately after an earthquake. Additionally, the MATLAB-based REM program supports direct import and efficient processing of TIF-format spatial data, with a single calculation taking only seconds, which is markedly more efficient than traditional methods. Validation using the Luding earthquake case shows that REM improves prediction accuracy by 13.706% compared to conventional methods. It accurately quantifies the synergistic impact of rainfall-earthquake coupling on slope stability and effectively identifies factor sensitivities. This research demonstrates that REM provides an efficient and practical tool for LSP in strong earthquake and heavy rainfall regions, playing a critical role in supporting disaster prevention planning and optimizing emergency decision-making.</p>