The present study uses nine machine learning (ML) methods to predict wave runup in an innovative and comprehensive methodology. Unlike previous investigations, which often limited the factors considered when applying ML methodologies to predict wave runup, this approach takes a holistic perspective. The analysis takes into account a comprehensive range of crucial coastal parameters, including the 2% exceedance value for runup, setup, total swash excursion, incident swash, infragravity swash, significant wave height, peak wave period, foreshore beach slope, and median sediment size. Model performance, interpretability, and practicality were assessed. The findings from this study showes that linear models, while valuable in many applications, proved insufficient in grasping the complexity of this dataset. On the other hand, we found that non-linear models are essential for achieving accurate wave runup predictions, underscoring their significance in the context of the research. Within the framework of this examination, it was found that wave runup is affected by median sediment size, significant wave height, and foreshore beach slope. Coastal engineers and managers can utilize these findings to design more resilient coastal structures and evaluate the risks posed by coastal hazards. To improve forecast accuracy, the research stressed feature selection and model complexity management. This research proves machine learning algorithms can predict wave runup, aiding coastal engineering and management. These models help build coastal infrastructure and predict coastal hazards. Graphical Abstract: [Figure not available: see fulltext.].