The advent of cloud computing has profoundly transformed the way computing resources are utilized, with the placement of virtual machines (VMs) being pivotal for enhancing resource allocation, improving energy efficiency, and maximizing overall system performance. This review paper provides traditional heuristic and metaheuristic approaches, as well as arising machine learning- based techniques, pressing their advantages and downsides. We explore an in- depth analysis of the rearmost algorithms and more sophisticated approaches for VM placement (VMP) in cloud environments, emphasizing their main objects, crucial characteristics, strengths, and limitations. The paper also discusses practical executions of optimized VMP and use case in areas like data center management, load balancing, edge computing, green cloud computing, energy efficiency, performance optimization and machine learning and AI workloads. Likewise, it addresses critical challenges like scalability, dynamic workloads, resource diversity, and heterogeneous resources, energy consumption and quality of service (QoS) conditions. Eventually, this review aims to establish a frame for future innovations in VMP. To achieve this, it incorporates the latest advancements and highlights areas that require further exploration. This approach improves the efficiency and sustainability of cloud computing environments.