The rapid growth of devices on the Internet of Things (IoT) has led to massive data generation, creating significant challenges in computational processing and storage. Multiaccess Edge Computing (MEC), combined with the introduction of 5G networks, has emerged as a promising solution to reduce latency and improve speed in data-intensive IoT applications such as smart agriculture, autonomous vehicles, augmented reality and telemedicine. MEC helps reduce latency by processing data closer to the source, making latency-sensitive applications more responsive. However, challenges such as efficient resource management, task allocation, and effective offloading mechanisms remain. This research proposes a K-means clustering algorithm-based approach to workload offloading in MEC systems, addressing application requirements, delay sensitivity, resource utilization, and system heterogeneity. The simulation results demonstrate that this method significantly reduces service time while optimizing the use of MEC resources. Furthermore, dynamic offloading decisions based on the computing capacity of the system and communication types can meet the evolving needs of modern IoT applications.