Here, we developed a comprehensive, manually curated single-cell multi-omics immune database with known cell labels in multi-species (scImmOmics, https://bio.liclab.net/scImmOmics), which aims to document a large number of available resources of high-quality single-cell immune multi-omics data, more broader and comprehensive cell/tissue types information. The current version of scImmRef documented over 2.9 million immune cells with cell type labels, involving 131 cell types and 47 tissues. These datasets have been manually curated from NCBI, HCA, EMBL-EBI, Github, Zendo and CellTypist, encompassing scRNA-seq, scTCR-seq, scBCR-seq, scATAC-seq, CITE-seq, ECCITE-seq and scCUT&Tag-pro. Notably, scImmOmics supplies rich regulatory information, such as clustering, gene expression, differentially expressed genes (DEGs) and their function annotation, differentiation trajectories, cell-to-cell communication, co-expression networks and their signature genes, protein expression, histone modification state, differentially chromatin accessibility regions and z-score of TFs. In addition, scImmRef offers analyses of clonal abundance, clonotype distribution across different cell types and immune responses to cytokines. In conclusion, scImmRef is a valuable and efficient resource that illuminates immune cellular heterogeneity, immune responses, and immune therapies at the single-cell level.