Social status is key to mental health and well-being. The nucleus accumbens (NAc) regulates motivated behaviors, including the establishment of social hierarchies. Our recent work underscored NAc mitochondrial energy metabolism as critical for social competition among rodents. However, the exact neurochemical and molecular underpinnings of the differences in NAc energy metabolism and the specific contribution of distinct cell types are yet to be determined. Magnetic resonance spectroscopy (MRS) allows the study of metabolites related to energy metabolism, neurotransmission, and cellular integrity in a non-invasive manner. Utilizing a gradient boosting machine learning algorithm, XGBoost, on 1H-MRS data obtained from male mice (N=81) living in tetrads, in which social status was assessed with the tube test, we first assessed which NAc metabolites are the strongest predictors of social status. Our machine learning model predicted social status with 77% accuracy. SHAP analyses indicated GABA, myo-inositol, glutathione, and lactate among the top-rated metabolites. Astrocytes are pivotal for GABA metabolism, release lactate which can be used as fuel by neurons, and are the main production site of glutathione. Subsequent analysis of the astrocyte translatome using RiboTag technology, highlighted differently expressed genes related to astrocyte morphology (microtubule organization) and energy metabolism (glycogen catabolism) between submissive and dominant mice. Targeted analysis revealed genes for GABA receptor subunits, transporters, and breakdown to be differently expressed in mice of different ranks. Future experiments targeting accumbal astrocytic functions are warranted to further uncover their causal role in social hierarchy formation.