Air ticket pricing is volatile, as various dynamic factors driving change alter demand levels, seasonal variations, booking time, and competitive pricing. It indeed becomes confusing and difficult for the airlines as well as passengers because the prices of tickets often change erratically. With such an understanding, the potential to accurately estimate ticket prices has grown to become a vital revenue maximization tool in airlines while similarly helping consumers make educated purchase decisions. Therefore, considering that such price variations may be quite challenging and unpredictable, the demand is gigantic for robust prediction models of prices that will give reliable estimations under a myriad of market conditions. Four widely used regression models to forecast air ticket prices are critically analyzed in this research work: Linear Regression, Polynomial Regression, Decision Tree Regression, and Gradient Boosting Regression. The models selected have already established presence in predictive analytics and differ in complexity, interpretability, and computational efficiency. Performance of each model is measured and compared according to many metrics: MAE, MSE, Recall, Precision, Accuracy and F1-Score. Our results show that Gradient Boosting Regression achieves the highest predictive accuracy, and it is by far the most reliable choice for a ticket price forecast.