This paper proposes a blending ensemble learning model to forecast crude oil prices, specifically focusing on Brent and WTI crude oil. The model combines various machine learning methods, including $k$-nearest neighbor regression, regression trees, linear regression, ridge regression, and support vector regression. The effectiveness of the proposed model is validated using Brent and WTI crude oil price data at different time series frequencies (daily, weekly, and monthly). The performance of the blending ensemble model is benchmarked against existing individual and ensemble learning methods, such as lasso regression, bagging lasso regression, boosting, random forest, and support vector regression. The results demonstrate that the proposed blending ensemble model outperforms existing models in terms of forecasting errors for both short- and medium-term horizons. The Diebold Mariano (DM) test further confirms the statistical significance of the model's superiority. The paper also discusses the advantages of the blending ensemble approach, including computational efficiency and the ability to capture diverse price patterns.This paper proposes a blending ensemble learning model to forecast crude oil prices, specifically focusing on Brent and WTI crude oil. The model combines various machine learning methods, including $k$-nearest neighbor regression, regression trees, linear regression, ridge regression, and support vector regression. The effectiveness of the proposed model is validated using Brent and WTI crude oil price data at different time series frequencies (daily, weekly, and monthly). The performance of the blending ensemble model is benchmarked against existing individual and ensemble learning methods, such as lasso regression, bagging lasso regression, boosting, random forest, and support vector regression. The results demonstrate that the proposed blending ensemble model outperforms existing models in terms of forecasting errors for both short- and medium-term horizons. The Diebold Mariano (DM) test further confirms the statistical significance of the model's superiority. The paper also discusses the advantages of the blending ensemble approach, including computational efficiency and the ability to capture diverse price patterns.