Gliomas are the most common tumor in the central nervous system in adults, with glioblastoma (GBM) representing the most malignant form, while low-grade glioma (LGG) is a less severe. The prognosis for glioma remains poor even after various treatments, such as chemotherapy and immunotherapy. Cuproptosis is a newly defined form of programmed cell death, distinct from ferroptosis and apoptosis, primarily caused by the accumulation of the copper within cells. In this study, we compared the difference between the expression of cuproptosis-related genes in GBM and LGG, respectively, and conducted further analysis on the enrichment pathways of the exclusive expressed cuproptosis-related mRNAs in GBM and LGG. We established two prediction models for survival status using xgboost and random forest algorithms and applied the ROSE algorithm to balance the dataset to improve model performance.