Abstract Rare cancers face major research challenges due to limited sample sizes and data scarcity. Existing pan-cancer databases mainly focus on common cancer types, while rare cancers often lack sufficient attention and systematic data collection. In addition, their high heterogeneity and the scarcity of studies on genomic features, immune environments, and drug responses lead to significant gaps in current databases and analytical tools. To address these limitations, we developed the Rare Cancer Explorer (RaCE), a dedicated database for integrated curation, analysis, and visualization of rare cancers. RaCE consolidates 5451 samples spanning 13 rare solid tumor types from 69 independent datasets, offering researchers a one-stop data analysis toolkit. The database provides one integrated dataset meta-analysis module and eight dedicated rare cancer functional analysis modules, including transcriptomics, immune infiltration, and immunotherapy response prediction, with a specialized focus on modeling gene effects and drug sensitivity in rare cancer cell lines. RaCE distinguishes itself through robust interactive functionalities, enabling users to seamlessly explore multi-layered insights from gene functions to therapeutic targets, thereby accelerating precision medicine and translational research for rare cancers. Compared to existing databases, RaCE demonstrates unique advantages in supporting rare cancer research through comprehensive data integration. The database is freely accessible at https://biospace.shinyapps.io/race/ or https://hiplot.com.cn/race/.