The advent of cloud computing has introduced a novel paradigm in user services, enabling on-demand access to information technology services, irrespective of time and location, with a pay-per-use model. Meeting user expectations necessitates cloud service providers to furnish an infrastructure that is efficient, reliable, and resilient to faults. Addressing the substantial volume of user requests, selecting cloud datacenters and load balancing strategies become pivotal for delivering real-time services economically and selecting appropriate datacenters and virtual machines. This research introduces DEThresh, combining a smart differential evolution-based datacenter selection method with a threshold-based load balancing mechanism aimed at enhancing resource management by focusing on cost and performance considerations. The differential evolution-based datacenter selection employs evolutionary operations to identify the most suitable DC for the userbase. Simultaneously, threshold-based load balancing ensures the reasonable distribution of tasks and optimal resource utilization, preventing task overload on virtual machines and underutilization due to task scarcity. Simulation results demonstrate that the proposed DEThresh algorithm exhibits superior response and data processing time compared to well-established popular algorithms currently in use, especially under varying userbase and datacenter conditions. Impressively, DEThresh achieves this even with fewer datacenters, ultimately reducing cloud service costs.