The growing complexity and scale of cyber threats pose significant challenges to the security and efficiency of high-speed industrial communication networks. As these networks play a crucial role in modern infrastructure, advanced cyber threat handling systems are essential to maintain their integrity. Traditional methods often struggle to effectively manage the dynamic nature of cyber threats and process high-dimensional data in real-time. To address these issues, this paper presents a novel Self-Adaptive Multi-Modal CyberThreat Intelligence (MMCTI) Model, designed to improve detection accuracy, scalability, and adaptability. The proposed model integrates multimodal datasets and advanced feature extraction techniques to enhance feature representation, enabling better prediction and handling of evolving cyber threats. This work introduces a dynamic approach that combines real-time prediction incorporating incremental threat learning over regular time-period, and threat mitigation alert generation, ensuring low-latency responses and the secure processing of high-volume network data. Extensive evaluations of the MMCTI model demonstrate a significant increase in threat detection accuracy and adaptability, positioning it as an effective solution for modern industrial communication networks, with notable improvements in handling diversity of cyber attacks.