<p>This study addresses artificial intelligence (AI)-powered decision-making for improving container port operations. The proposed model integrates quantum computing (QC) with an improved soft actor-critic (SAC) within a deep reinforcement learning (DRL) framework. Variational quantum circuits (VQCs) enhance the action selection for the container premarshalling problem (CPMP), ensuring container terminal efficiency and productivity. In CPMP, significant challenges include container rearrangement, resource allocation, and crane movement optimization by eliminating blockages. Computational experiments validated the hybrid quantum-classical algorithm within deep learning paradigms for practical port operations, reducing container relocations and crane operation times. Sensitivity and ablation studies highlighted the critical role of quantum circuits, heuristics, and entropy regularization, showing a quantum-classical synergistic impact on model performance. This innovative method demonstrates the potential of quantum-enhanced deep learning strategies, to offer more effective solutions for premarshalling with crane time objectives, than traditional methods which face several challenges, improve multi-objective optimization, and extend the foundations for future quantum computing applications.</p>