The mental stress among people, especially the modern working class living in the concrete jungles, has been growing at a drastic rate. And the major reason why their stress problem can’t be solved easily is that they don’t know about the mental stress upon them. It is often seen that student who live in a Nuclear Family have a higher chance of getting stressed than those who live in a Joint family; maybe this might be False in some cases, our idea is to measure their stress level using our machine learning algorithms. Our results will help them to recognize the ocean of threats they have been drowning in. Thus, the f1 score , confusion matrix, recall and other types of measuring techniques were added that gave us the error and specificity values, which helped identify the best accuracy model among some applied algorithms as the XG Boost, Random Forest, Decision Trees and Furthermore, and deep learning models like Simple Feedforward Neural Network (FNN), the specificity parameter revealed that the algorithms were also sensitive to some negative results.