<p>Precise estimation of the elastic modulus of recycled aggregate concrete (RAC) is vital for the development of sustainable construction works. In this article, a hybrid modeling strategy integrates the utilization of both Support Vector Regression (SVR) and Adaptive Boosting Regression (AdaBoost) with two metaheuristic optimizers, i.e., Political Optimizer (PO) and Chef-Based Optimization Algorithm (CBOA), with the aim of improving predictive performance. A range of published experimental investigations provided the extensive dataset that was used; the appropriate input features were determined based on the F-statistic method. RMSE, R-squared, Mean Absolute Relative Error (MARE), n10-index, and Mean Normalized Bias (MNB) parameters were used for the evaluation of the models. The SVR-Political Optimizer (SVPO) model performed the most effectively among the six built models with an R-squared value of 0.976 and a lowest RMSE of 1.233. These findings show the model’s potential in identifying intricate relationships within diverse RAC data. The outcomes set out the possible role of coupling machine learning and optimization methods for the accurate prediction of RAC properties with a practical tool available for quality control, mix design, and sustainable structural engineering.</p>