<p>The integration of digitalization and decarbonization presents a new opportunity to advance the United Nations Sustainable Development Goal 13, which aims to combat climate change and its impacts. In this context, China has launched the Carbon Generalized System of Preferences (CGSP) as a digital initiative supporting Sustainable Development Goal 13. By leveraging big data technology to track and account for individuals’ carbon footprints, the CGSP functions as a low-carbon behavior reward system that operates through carbon coin income (carbon reduction rewards) and expenditure (product redemption) within individuals’ carbon accounts. However, the CGSP faces dilemmas such as low user engagement, a high proportion of “zombie” users, and insufficient carbon reduction in practice. To address these challenges, we establish a theoretical model from the perspective of mental accounting. Through one experimental study and one empirical study, we reveal the mental accounting mechanism of CGSP users after engaging in low-carbon behaviors and its impact on perceived value and continued participation. The results show: (1) There is a significant mental accounting effect in users’ behavioral decisions. The carbon coins earned from different low-carbon behaviors are categorized into functional, emotional, and social accounts, leading to differences in income perception; (2) Carbon coins in these three mental accounts exhibit significant differences in product preference during redemption, resulting in variations in expenditure utility; (3) The perceived value after redeeming carbon coins differs depending on the attributes of the products. The redemption of practical goods and public welfare goods significantly increase users’ perceived value, which in turn stimulates new low-carbon behaviors, while other types of carbon coin expenditures fail to provide adequate positive incentives for sustained behavior. These findings provide important theoretical and practical contributions.</p>