The interaction between cations and delocalized electronic clouds (the cation-π interaction) occupies a very important place within non-binding interactions. Its presence has long been recognized as fundamental for both, the structure and function, of proteins and other important biological molecules. Rechargeable batteries and fuel cells industries of are also interested in cation-π interaction and the use of graphene and similar carbon allotropes are investigated as promising alternatives in their technological applications. Reliable and practically applicable theoretical models of cation-π interaction are needed for guiding these researches. In this work, the interaction of cations (Li⁺, Na⁺, K⁺, ammonium and guanidinium) with graphene fragments (from benzene to circumcoronene) is modeled using DFT level of theory. Linear scans (TPSS+D3/Def2TZVPP) that follow trajectories perpendicular to the central ring of the graphene fragments allow the location of the distance at which the strongest interaction takes place. Using the geometry of the minima, the interaction energy is decomposed in physically meaningful contributions using a SAPT(DFT) method (PBE0/Def2TZVP). It is observed that benzene complexes systematically deviate from the trend followed by complexes with larger fragments, so this system does not constitute a good model for the study of cation-π interaction in graphenes or other large conjugate molecules. While induction is the main contribution in complexes with Li⁺ and Na⁺, the stability of most of the complexes investigated depends on a balanced combination of the three contributions: electrostatic, induction and dispersion. Following the tendencies observed with organic fragments with an increasing number of conjugate rings, the results can be extrapolated to extended π systems as graphene.