Higher-order chromatin topology is essential for cellular function, yet mapping simultaneous multi-way chromatin contacts at high resolution in single cells remains a significant challenge. Existing single-cell three-dimensional (3D) genome mapping methods are constrained by sparse data recovery. Here, we present scOmniMicro-C, a method that integrates micrococcal nuclease (MNase) digestion with a combination of short- and long-read sequencing to map single-cell 3D genomes. By preserving more comprehensive contact information, scOmniMicro-C achieves an unprecedented average of 6.2 million contacts per cell—substantially surpassing the efficiency of current assays employing a single sequencing strategy. This data density enables the reconstruction of diploid human 3D genomes at 1-kb resolution and haploid yeast genomes at 200-bp resolution. Using this approach, we resolved the spatial clustering of active regulatory elements from high-resolution 3D reconstructions, identified specialized "Nucleating" and "Hub-Core" enhancers within super-enhancers through conditional probability analysis, and revealed cooperative interactions among enhancers regulating the same gene. We further applied scOmniMicro-C to characterize centromere clustering, nucleolar subcompartments, and cohesin-mediated chromatin loop hubs in Saccharomyces cerevisiae. scOmniMicro-C provides a robust, high-efficiency framework for mapping chromatin architectures and intricate regulatory hubs in single cells.