Current breast cancer classification methods, particularly immunohistochemistry and PAM50, face challenges in accurately characterizing the HER2-low subtype, a therapeutically relevant entity with distinct biological features. This notable gap can lead to misclassification, resulting in inappropriate treatment decisions and suboptimal patient outcomes. Leveraging RNA-seq and machine-learning algorithms, we developed the Breast Cancer Classifier (BCC), a unique transcriptomic classifier for more precise breast cancer subtyping, specifically by delineating and incorporating HER2-low as a distinct subtype. BCC also redefined the PAM50 Normal subtype into other subtypes, disputing its classification as a unique molecular group. Our statistical analysis not only confirmed the reproducibility and accuracy of BCC, but also revealed similarities in prognostic characteristics between the HER2-low and Basal subtypes. Addressing this gap in breast cancer classification is clinically significant because it not only improves treatment stratification, but also uncovers novel molecular and immunohistochemical features associated with the HER2-low and HER2-high subtypes, thereby advancing our understanding of breast cancer heterogeneity and providing guidance in precision oncology.