This study evaluates the potential of machine learning algorithms for early-stage diabetes prediction. A dataset containing demographic information, medical history, and lab results was analyzed using Logistic Regression and Random Forest Classifier. The results showed that Random Forest algorithms were able to accurately predict diabetes at an early stage with high accuracy. The best-performing algorithm was found to be the Random Forest Classifier, with an accuracy of 98.0%. These findings suggest that machine-learning algorithms hold great promise for improving diabetes diagnosis and management. The results of this study provide valuable insights for future research in this area and may help inform the development of more effective and efficient screening and treatment strategies for diabetes.