The rapid population growth and urbanization in Jakarta pose significant challenges to the provision of efficient public transportation, particularly for Transjakarta, which often experiences fluctuating passenger volumes that complicate capacity management and operational efficiency. This study aims to model and predict Transjakarta passenger volumes using regression methods within machine learning algorithms, by comparing three models: Linear Regression, Random Forest Regression, and Gradient Boosted Trees Regression. The dataset consists of historical passenger records from routes S21 (Ciputat–CSW/Tosari) and S22 (Ciputat–Kampung Rambutan) covering the period from January 2022 to March 2025. The data were processed through several stages, including preprocessing, categorical variable transformation, train-test splitting, and model evaluation using Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and the coefficient of determination (R²). The results show that Gradient Boosted Trees Regression achieved the best predictive performance with an R² of 0.73 and an average error of approximately 22,000 passengers, outperforming Linear Regression (R² = 0.65) and Random Forest Regression (R² = 0.63). These findings highlight that ensemble boosting is more effective in capturing non-linear patterns in passenger data, making it the most suitable predictive model to support operational planning, fleet efficiency, and the development of adaptive and sustainable public transportation policies.