Copper (Cu) single-atom catalysts (SACs) exhibit great potential for generating multicarbon (C 2+ ) products, but the intrinsic activity of single-atom Cu (Cu 1 ) under realistic conditions remains controversial. Herein, we perform extensive calculations with explicit solvation to investigate the underlying mechanism of Cu SACs, disclosing the absence of C 2+ activity in Cu 1 sites regardless of the different substrates. The original Cu 1 sites (first taking Cu 1 stably anchored on carbon nitride as an example) cannot facilitate *CO hydrogenation and CO-CO coupling due to the lack of active sites nearby, and they are unstable under operation, causing leaching and aggregation to form small Cu clusters. The derived Cu clusters composed of at least three Cu atoms can efficiently promote CO-CO coupling, as revealed by kinetic analyses. We extend the modeling to other typical Cu SACs and reveal that all of the Cu 1 sites are inactive, while the C 2+ performance of the derived Cu-cluster catalysts is substrate-dependent. This study offers mechanistic insights into Cu SACs and provides practical guidance for their rational optimization.