Cloud computing services refer to the on-demand provision of computer resources and services over the internet. Numerous resources are available from cloud service providers (CSPs). Infrastructure as a service (IaaS) is a cloud computing service that enables the sharing of computer resources over the web. One of the key challenges in cloud scheduling is the efficient allocation of these resources. Recently, several swarm-intelligence (SI) scheduling techniques have been adopted. In this study, a two-phase scheduling model known as the clustered grey wolf optimization (CGWO) algorithm is proposed. During the first phase, the task splitting agglomerative clustering (TSAC) algorithm classifies jobs based on their deadlines, while the advanced grey wolf optimization (AGWO) algorithm handles resource allocation. The CloudSim simulation results demonstrate that the CGWO framework outperforms currently used algorithms, including genetic algorithm (GA), particle swarm optimization (PSO), salp swarm algorithm (SSA), and standard grey wolf optimization (GWO). The evaluation considers factors such as makespan, resource utilization, cost, throughput, convergence speed, and others when comparing various cloud scheduling algorithms. The suggested model incorporates a clustering mechanism to alter the traditional first-in, first-out (FIFO) structure of job execution. This study reveals that GA and SSA are excellent choices, particularly for lower and intermediate task counts, if the primary goal is to reduce makespan. If effective resource utilization and throughput are top priorities, SSA and CGWO appear to be promising options. The improvement rate of SSA over CGWO in terms of makespan is approximately 0.135%. Regarding resource utilization, CGWO has shown an improvement rate of 8.228%, 4.88%, and 0.93% compared to GA, GWO, and PSO, respectively. CGWO's rate of resource utilization improvement is 1.28% lower than that of SSA.