The multi-scale dynamics of oceanic processes and the complex propagation of acoustic waves are fundamental challenges in marine sciences and operations. Recent computing advances enable such multiresolution ocean and acoustic modeling, but a fully integrated system for sustained coupled predictions and Bayesian data assimilation remains needed. In this study, we integrate the MSEAS Primitive Equation (PE) ocean modeling system and the MSEAS acoustic Parabolic Equation (ParEq) solver, enabling real-time coupled ocean and acoustic predictions. Realistic applications in Massachusetts Bay, the Norwegian Sea, the western Mediterranean Sea, and the New York Bight are used to demonstrate capabilities and validate predictions in diverse shallow and deep-water environments. Results provide the foundation for an end-to-end system for coupled ocean-acoustic probabilistic modeling, Bayesian inversion, and learning.