Tropical cyclones are prominent sources of atmospheric gravity waves due to their intense and organized deep convection, yet the sensitivity of cyclone-induced gravity waves to model physics choices remains poorly constrained. Here, we examine how different physical parameterizations in the Weather Research and Forecasting (WRF) model influence the generation and characteristics of stratospheric gravity waves during the well-observed case of Typhoon Soudelor (2015). A suite of high-resolution simulations of the tropical cyclone employing multiple microphysics, planetary boundary layer, and cumulus parameterizations is analyzed. Gravity wave diagnostics are derived from vertical velocity variability and a localized S-transform spectral analysis. Although the simulations produce similar tropical cyclone tracks and intensities, they generate markedly different gravity wave responses. Differences in microphysics and boundary layer schemes lead to systematic variations in gravity wave amplitudes and spectral characteristics, while the inclusion of a cumulus parameterization consistently weakens deep convection and gravity wave amplitudes. These differences occur despite comparable background wind and stability conditions, demonstrating that gravity wave sensitivity is largely controlled by differences in diabatic forcing among the simulations. Our results highlight the critical role of model physics in shaping gravity wave generation and stratospheric wave characteristics and demonstrate the importance of carefully evaluating diabatic forcing in mesoscale modeling studies.