This research is devoted to the study of stroke since it is a terrible medical condition that has enormous monetary and societal consequences. Stroke is becoming more common as a result of an ageing population, thus scientists are developing automated algorithms to predict these events in the hopes of saving lives via early intervention. Examining several machine learning classifiers and proposing explainable approaches like Shapley additive explanation (SHAP) and local interpretable modelling agnostic explanation (LIME), this research aimed to better understand how medical models make decisions. The findings showed that more complex models achieved higher levels of accuracy; for instance, the best model obtained about 91% accuracy, while the remaining models achieved accuracy that ranged from 83 to 91%. This paradigm integrates local as well as worldwide explainable techniques to enhance stroke care and treatment. It provides understanding of intricate models.