In this research paper, we explored the efficacy of various machine learning algorithms in predicting domain validity, an essential task for cyber security and online safety. In response, we introduce a predictive model that leverages Machine Learning and Deep Learning approaches. Our dataset comprises 95910 instances, encompassing 12 features including the class label. Notably, this dataset comprises 55914 invalid domains and 39996 valid domains. This study evaluates the performance of seven different classifiers: K-Nearest Neighbor, Random Forest, Support Vector Machine, Naïve Bayes, Neural Network, Logistics Regression, and XG Boost. We systematically evaluated 26 models’ performance with different scenarios. Among them XG Boost emerged as the most effective, boasting an accuracy 95.50%, Precision 96.10%, Recall 96.25%, and F1-Score 96.18%. This model provides the best balance of high precision, recall, and F1-Score, alongside high accuracy, making it the most robust choice for our classification problem.