Diabetes is a disease that should be detected early because it cannot be cured permanently. Using data mining and machine learning (ML) techniques, diabetes has been predicted. In order to successfully forecast diabetes, machine learning techniques are frequently applied. The Pima Indian Diabetes (PID) dataset, which contains data on seven sixty-eight patients and the 9 unique traits that are unique to each of them. Diabetes has been predicted using seven distinct machine-learning techniques. Two of the techniques utilized for diabetes Mellitus are Logit model and Support Vector Machine (SVM). When implemented with Weka, they achieve approximately 77% accuracy, while when implemented with Python, they achieve approximately 83% accuracy.