Accurate prediction of total operation time is crucial for optimizing manufacturing processes. This study investigates the application and evaluation of various regression machine learning (ML) models for predicting total operation time in longitudinal multi-pass turning of a part with complex geometry made of S235 construction steel. Using a dataset generated through simulations in FeatureCAM software, the performance of Gradient Boosting Regressor (GBR), Random Forest Regressor (RFR), and K-Nearest Neighbors (KNN) Regressor was assessed. The GBR model emerged as the most effective, achieving near-perfect predictive accuracy with minimal error metrics, attributed to its iterative boosting mechanism. The RFR also demonstrated high accuracy, though with slightly higher errors, while the KNN underperformed significantly. The findings highlight the potential of advanced ML models to enhance manufacturing efficiency by providing reliable estimations of total operation time. This research contributes to the field by offering practical insights into the integration of ML models in manufacturing, underscoring their transformative potential in improving production efficiency and reducing operational costs.