Text classification, mainly in the realm of online interaction of people in general through text, plays a vital role in maintaining safety over any social platform. Especially, in this age where it is easy to go online and type without filter. This research focuses on a comparative study of two simple yet prominent classifiers - Logistic Regression (LR) and Naive Bayes - to classify provided text as either toxic or non toxic. This research identifies potential improvements through iterative training and data processing enhancements.