Injection molding remains the dominant process for high-volume plastic production, but its strong sensitivity to process parameters, material behaviour, and cooling efficiency makes traditional trial-and-error optimization costly and slow. Moldex3D, a dedicated 3D CAE package for polymer processing, enables detailed simulation of filling, packing, cooling, and warpage, allowing defects such as short shot, weld lines, air traps, sink marks, voids, and deformation to be predicted before tool manufacture. By combining non-Newtonian flow models, temperature-dependent material data, and realistic mold and cooling layouts, Moldex3D helps designers and process engineers optimize gate locations, runner balance, packing profiles, and conventional or conformal cooling channel designs. This review consolidates recent Moldex3D-based research across automotive, consumer, electronic, and medical applications, with emphasis on flow and shrinkage analysis, warpage prediction in fibre-reinforced parts, and use in multi-cavity and thin-walled molds. Advanced and hybrid workflows are also examined, including integration with CAD/CAE platforms, topology and DOE-based optimization, and transfer of fibre orientation and residual stress fields into structural FEA. Key limitations are identified in material modeling (PVT, viscosity, fibre orientation), mesh and computation cost for full 3D models, and incomplete coupling to structural durability analysis and real machine behaviour. Finally, the review highlights future opportunities for AI-assisted optimization, cloud-based simulation, and digital twin integration, positioning Moldex3D as a core enabler of simulation-driven, intelligent injection molding.