Future sea level rise under global warming poses serious risks of extreme sea level events in coastal regions worldwide. Numerous state-of-the-art climate models, with their relatively coarse horizontal resolution, may not adequately resolve coastal wave dynamics, leading to uncertainties in coastal sea level variability representation. This study compared eddy-resolving and non-eddying ocean models in reproducing sea level variability, focusing on the probability distribution along the western coast of India. The eddy-resolving model can simulate intraseasonal sea level variations associated with coastal waves driven by equatorial wind anomalies. The non-eddying model fails to capture over 81% of observed extreme sea level events, as shown in the probability distribution for intraseasonal time series. Although capable of simulating Indian Ocean Dipole-related low-frequency sea level anomalies, the non-eddying model does not replicate their connection to intraseasonal extreme events. The results suggest that climate model projections may underestimate future changes in extreme sea level events.