Real-time optimization of virtual reality (VR) models in dynamic systems demands adaptive decision-making frameworks capable of reconciling conflicting objectives such as computational efficiency, latency reduction, and user experience fidelity. While traditional optimization techniques often struggle with the non-linear, high-dimensional, and time-sensitive nature of VR environments, this paper introduces a novel hybrid framework that synergizes Goal Programming (GP) and Neural Network Artificial Intelligence (NN-AI) to address these challenges. The proposed methodology leverages GP to formalize multi-objective decision-making under constraints, while a dynamically trained neural network predicts and prioritizes system states in real time, enabling context-aware adjustments to VR model parameters. By integrating GP’s structured optimization with NN-AI’s predictive adaptability, the framework achieves Pareto-optimal solutions that balance competing objectives across fluctuating operational conditions. The study validates the framework through a series of simulated and real-world VR scenarios, including immersive gaming and industrial training systems, where dynamic variables such as user interactions, environmental complexity, and hardware limitations are present. Results demonstrate a 22–35% improvement in rendering efficiency and a 40% reduction in latency compared to conventional single-objective optimization approaches, without compromising visual quality. Furthermore, the system exhibits robust generalization capabilities, adapting to unseen scenarios within 5–10 iterations. This research bridges a critical gap in real-time multi-objective optimization for VR, offering a scalable, AI-driven solution for industries reliant on immersive technologies. The framework’s ability to harmonize human-centric objectives with computational constraints positions it as a transformative tool for next-generation dynamic systems in entertainment, healthcare, and Industry 4.0 applications.