This study investigated the effectiveness of a differentiated instructional model grounded in STEAM principles, augmented by an interactive PILAR media, on the memorization of Surah At-Tin among fourth-grade elementary students. A true experimental pretest–posttest control-group design was employed, involving an experimental cohort (n=6) and a control cohort (n=9). Both groups completed a baseline assessment of Qur'anic memorization before undergoing four instructional sessions; the experimental group received STEAM-based differentiated activities and digital media support, whereas the control group experienced conventional lecture-based instruction. Posttest results revealed that the experimental group achieved a mean score of 91.67 (SD=8.54), compared to 69.44 (SD=15.32) in the control group. Normalized gain analysis indicated a high gain (g=0.90) for the experimental cohort and a moderate gain (g=0.64) for the control cohort. These findings demonstrated that aligning pedagogical strategies with individual learning preferences, integrating multimodal STEAM tasks, and leveraging interactive technology significantly enhanced both the quantity and accuracy of Qur'anic memorization. The study concluded that a differentiated STEAM-based approach, supported by PILAR media, constituted a superior method for optimizing primary-level Qur'anic memorization and recommended its broader application and longitudinal evaluation.