<p>The performance of chemical vapor deposition (CVD) systems is critically governed by reactor geometry, flow behavior, and thermal gradients, particularly in vertical configurations tailored for high-aspect-ratio thin-film fabrication. This study presents a comprehensive, multi-scale optimization framework that integrates high-fidelity computational fluid dynamics (CFD) simulations, porous media augmentation, and surrogate modeling to enhance deposition rate and film uniformity. A vertical CVD reactor was first modeled and rigorously validated against benchmark data to ensure accuracy in capturing transport and surface reaction phenomena. To mitigate recirculation zones and improve gas-phase distribution, porous media were strategically embedded within the reactor. This modification resulted in improved thermal and species uniformity, directly influencing film quality. Following reactor enhancement, a systematic parametric study was conducted by varying susceptor temperature and inlet gas velocity. The resulting dataset was used to construct a second-order polynomial regression model, serving as a surrogate for rapid response surface analysis and optimization. To identify the optimal operating conditions, gray relational analysis (GRA) was employed as a robust multi-response optimization technique, effectively reducing the experimental burden while capturing key process interactions. The model demonstrated excellent predictive accuracy and provided valuable insights into process sensitivities. This integrated CFD–surrogate modeling–GRA approach offers a scalable and computationally efficient pathway for optimizing complex CVD systems. The findings underscore the potential of porous medium engineering and data-driven optimization in advancing next-generation thin-film deposition technologies.</p>