Load balancing methods ensure efficient resource allocation and timely system response in cloud computing. Algorithms of load balancing help mitigate constraints, adapt to fluctuating workloads, and maintain high availability. This research explores workflow scheduling and load balancing in cloud computing, focusing on optimizing resource use, reducing job completion, and enhancing user satisfaction. This study presents a comprehensive analysis of the performance of various computational models - Multi Swarm, Multi Objective, Modified Artificial Bee Colony (MABC), and Proposed - in RNA-based computations, with a focus on cost, time span, and energy parameters. The analysis is conducted under varying computational loads, represented by an increasing number of Virtual Machines (VMs), ranging from 2 to 20. The study encompasses both general RNA analysis and specific AUDIO-based scenarios, providing a holistic view of model performance across different applications. Key findings reveal that the MABC method generally offers superior cost efficiency across increasing VM counts, demonstrating its effectiveness in resource allocation and cost management. In terms of time span, while MABC excels in general RNA analysis, the Multi Objective model stands out in AUDIO-based scenarios, especially at higher VM counts. Regarding energy consumption, although MABC performs well, the Multi Objective model exhibits better energy efficiency in certain AUDIO-based instances.The study identifies 20 VMs as the optimal count for balancing computational resources with the efficiency and performance of the models. This finding highlights the importance of model selection based on specific task requirements and the nature of the computational task. The insights provided by this study are crucial for optimizing RNA-based computational processes, guiding the selection of appropriate models and resource allocation to achieve efficient, cost-effective, and sustainable outcomes in various computational settings.