Selections of metallic cathode materials and modulations of metal–sulfur bonding strength are crucial for sulfur immobilization in development of high-performance lithium–sulfur (Li–S) batteries with low cost. By combining theoretical calculations and experiments, herein we reveal the relationship between intrinsic electronic structure and metal–S bonding strength, which links to energy density and durability of Li–S batteries. Through first-principles calculations, we simulate sulfur clusters (S1, S2, S4, and S8) immobilization on metal (Cu, Ni, and Sn) slab surfaces with and without graphene substrate. For sulfur clusters, the metal–Sx (x = 1, 2, 4, and 8) bonding strength is in the sequence of Ni > Cu > Sn without graphene substrate. Nevertheless, the sequence changes (Ni > Sn > Cu) in the presence of graphene substrate due to different amounts of charge transfer between these metal clusters and graphene. Guided by these theoretical results, metal (Cu, Ni, Sn)/graphene (G) composites are synthesized and subsequently integrated into the cathode of Li–S batteries. Among these metal/G systems, the sulfur cathode with Ni/G composites demonstrates remarkable electrochemical performance, i.e., a discharge capacity of >830 mAh g–1 over 500 cycles with an average Coulombic efficiency close to 100%. These findings shed light on theoretical calculations providing insights into the electrode design of Li–S batteries.