In today’s rapidly growing real estate market, especially in urban cities like Bengaluru, there is a growing need for accurate housing price predictions to support buyers, sellers, and developers. This research addresses this requirement by comparing six machine learning techniques—Linear Regression, Decision Tree, Random Forest, Gradient Boosting, XGBoost, and Neural Network—applied to structured property data. Among these, XGBoost emerged as the most effective technique, achieving the lowest Mean Absolute Error (0.57), Root Mean Squared Error (0.92), and the highest R2 score (0.90), making it the most accurate model for predicting house prices. In contrast, Linear Regression performed the worst, with an R2 score of 0.66, highlighting its limitations in handling the complexity of the dataset. The importance of this research lies in its practical application, as it not only evaluates model performance but also implements these models into a full-stack web application, making price predictions accessible to end-users through a modern interface.