Cloud computing has grown rapidly in recent years, mainly due to the sharp increase in data transferred over the internet. This growth makes load balancing a key part of cloud systems, as it helps distribute user requests across servers to maintain performance, prevent overload, and ensure a smooth user experience. Despite its importance, managing server resources and keeping workloads balanced over time remains a major challenge in cloud environments. This paper introduces a novel Score-Based Dynamic Load Balancer (SBDLB) that allocates workloads to virtual machines based on real-time performance metrics. The objective is to enhance resource utilization and overall system efficiency. The method was thoroughly tested using the CloudSim 7G platform, comparing its performance against the throttled load balancing strategy. Evaluations were conducted across a variety of workloads and scenarios, demonstrating the SBDLB's ability to adapt dynamically to workload fluctuations while optimizing resource usage. The proposed method outperformed the throttled strategy, improving average response times by 34% and 37% in different scenarios. It also reduced data center processing times by an average of 13%. Over a 24-hour simulation, the method decreased operational costs by 15%, promoting a more energy-efficient and sustainable cloud infrastructure through reduced energy consumption.