Gastric cancer is characterized by high heterogeneity, with its complex microenvironment and intercellular communications playing critical roles in disease progression and treatment responses. In this study, we utilized single-cell sequencing to dissect the intricate landscape of gastric cancer, identifying diverse cell populations and their interactions. We focused on the role of β-hydroxybutyrylation (Kbhb)-associated genes and their impact on the tumor microenvironment. By analyzing 189,700 single-cell profiles, we identified four distinct malignant epithelial cell subpopulations characterized by unique gene expression patterns. Among these, 20 β-hydroxybutyrylation (Kbhb)-associated genes were identified, including key genes such as MRPL13, LDHB, COX6C, FABP5, and RPS13, which were significantly associated with immune infiltration and tumor microenvironment remodeling. Hierarchical clustering based on these genes classified gastric cancer patients into two subgroups with distinct prognostic outcomes. Patients in the high-risk subgroup exhibited increased expression of pro-tumor genes and reduced immune infiltration, correlating with poorer survival. We further constructed a robust risk scoring model incorporating these genes, achieving AUC values of 0.72, 0.69, and 0.66 for predicting 1-, 3-, and 5-year survival in the TCGA dataset. These findings underscore the prognostic value of Kbhb-associated genes and their potential as therapeutic targets. This study not only provides insights into the molecular underpinnings of gastric cancer but also offers potential biomarkers for patient stratification and targets for therapeutic intervention.