The most common chronic diseases, diabetes mellitus, affect millions of people annually throughout the world. In order to lower the long-term health risk of diabetes, such as heart disease, kidney failure, and nerve damage, early detection and management are essential. The order to predict the risk of diabetes uses actual clinical data; this study presents a hybrid model that combines the Random Forest (RF) and Logistic Regression (LR) algorithms. Increase accuracy and interpretability, model also use Recursive Feature Elimination (RFE) to identify the most significant predictive features.PIMA Indian Diabetes dataset, along with World Health Organization (WHO) global health data, was used to train and validate the suggested model. The hybrid LR–RF approach obtained an accuracy of 89.2%, based on the findings and outperformed the individual model with a ROC-AUC score of 0.91. This model method shows how data-driven and interpretable artificial intelligence can help with clinical decision-making and provide patients and healthcare providers with trustworthy diagnostic tools.