In Cloud Computing (CC), load balancing tasks remain an essential problem of spreading resources from a data center to ensure that each Virtual Machine (VM) has a balanced load to achieve maximum utilization of its capabilities. In the CC world, load balancing is a Non-Polynomial (NP) problem solved with metaheuristic algorithms. A new Quasi Oppositional Dragonfly Algorithm for Load Balancing (QODA-LB) was developed to achieve the optimal resource scheduling in a CC setting. The proposed QODA-LB algorithm uses three variables to compute an objective function: run time, running cost, and load. The QODA-LB algorithm assigns tasks to VM based on its potential and the derivative objective function. Also, the QODA-LB algorithm uses the principle of Quasi-Oppositional Based Learning (QOBL) to increase the standard Dragonfly Algorithm's (DA) convergence rate. A comprehensive series of experiments were conducted, and the findings were analyzed in a variety of ways to ensure the efficient performance increased by the QODA-LB algorithm. The simulation's results demonstrated optimum load balancing efficiency and outperformed the leading approaches.