BackgroundThe involvement of glycolysis in carcinogenesis and the tumor microenvironment is being increasingly supported by the available data. The aim of this work was to develop a triple-negative breast cancer predictive model based on glycolysis. MethodsGlycolysis mediated pattern clusters were created using the R “ConsensusClusterPlus” package. The variations in the tumor microenvironment between the pattern clusters were examined using the R “GSVA”, “ESTIMATE”, and “CIBERSORT” package. The risk score and nomogram were established to assess clinical outcomes of patients. ResultsSubstantial differences were observed in the immunological landscape between the glycolysis-mediated pattern clusters. When it came to predicting survival and immunotherapy response, the developed risk score showed promising predictive power. Nomogram was designed to be used in therapeutic settings due to its remarkable predictive accuracy. ConclusionsThe immune microenvironment varied among cases of triple-negative breast cancer. The nomogram and the risk score based on glycolysis could both be used to help create more effective treatments.