Network analysis is pivotal for understanding intricate structures, yet managing large and complex networks presents a formidable challenge for existing algorithms. In this research, we introduce MNC-MDN, a pioneering Multi-level Network Coarsening framework that integrates density estimation with hierarchical strategies. This innovative approach adeptly identifies high-density components and hierarchically coarsens networks, effectively reducing complexity while preserving crucial elements such as most dominant neighbors and influential edges. Experimental findings on diverse real-world datasets highlight that MNC-MDN excels in achieving a fine balance between coarsening accuracy and computational efficiency, particularly excelling in accuracy for expansive networks. Furthermore, it provides detailed insights into community structures across multiple hierarchical levels.