The goal of this work is to develop a model that can accurately detect malicious URLs. The model should be able to classify raw URLs into different types such as benign (which means safe URLs), or malicious URLs (which includes phishing URLs, malware URLs, or defacement URLs). The work addresses the limitations of traditional techniques such as URL blacklisting or signature blacklisting, which offer only “after-the-fact” solutions. The model can generalize to new URLs and improve the timeliness and abstraction of malicious URL detection methods. The work involves creating a feature set representation for identifying malicious URLs and measuring the model’s ability to predict the maliciousness of URLs. The dataset used for the work consists of many URLs categorized as malicious or benign. The work involves constructing the efficient XGBoost algorithm with fine tuning of the hyper parameters and plotting average feature importance to understand which features are important in predicting malicious URLs.