Phishing is a type of fraud where the attacker attempts to get private information and as the technology progress, Phishing attacks have progressed and enhanced their techniques. This work explores advanced machine learning techniques for detecting URL phishing attacks, a major cybersecurity threat. It highlights the importance of machine learning in enhancing URLs security. This paper reviews existing literature and then details methodologies for data collection, preprocessing, feature selection, and model training. It presents and analyzes results of three machine learning models, including XGBoost, SVM and Random Forest using wide range of evaluation metrics such as Accuracy, Precision, Recall, F1-score, Area Under Curve (AUC). The results obtained from our experiment, revealed that the Random Forest model achieved the highest scores in all validation tests, with an accuracy of 0,91, followed by the XGBoost with an accuracy rate of 0,88.