We analyze the impact of smartphone usage on multidimensional poverty reduction. Utilizing the A-F approach, we compute a multidimensional poverty index based on five dimensions: education, health, income, living standard, and labor force. This index categorizes multidimensional poverty into three levels: Vulnerable Multidimensional Poverty Index (VMPI), General Multidimensional Poverty Index (GMPI), and Extreme Multidimensional Poverty Index (EMPI), following MPI criteria. Furthermore, we investigate the mediating role of social capital in the smartphone-multidimensional poverty relationship through a mediating effects analysis. We used the survey data of 382 sample out-of-poverty rural households in Jiangxi, China, in 2020. Our results indicated that: (1) Education (37.80%), labor force (29.7%), and health (20.40%) were identified as the primary contributors to multidimensional poverty. (2) Increasing deprivation categories correlated with declining multidimensional poverty index, following an inverted U-shaped pattern. (3) Smartphone usage significantly reduced VMPI (57.6%), GMPI (52.6%), and EMPI (5%). (4) Social capital fully mediated EMPI reduction through smartphones (91.67%), and partially mediated VMPI (14.09%) and GMPI (20.84%) reduction. These insights inform targeted policy formulation for rural multidimensional poverty reduction.