In this study, density functional theory was used to optimize five catalyst models consisting of one Cu 38 nanoparticle and four Zn-, Pd-, Ag-, and Audoped Cu 37 -X nanoparticles. The energy gap value of the Cu 38 nanoparticle catalyst was calculated as 15.3 kcal mol −1 , and the energy gap values of Cu 37 -X (X: Au, Ag, Zn, Pd) were 15.0 kcal mol −1 , 15.2, 11.2, and 12.9 kcal mol −1 , respectively. All of these catalysts are photoactive in the visible-light range and can be used as good photocatalysts. The photocatalytic mechanism of CO 2 reduction on these optimized catalysts has been investigated in detail. Two possible reaction pathways for photocatalytic reduction of CO 2 to CO and four possible reaction pathways for the further reduction of CO to CH 3 OH were investigated. In the first stage of catalytic reduction, i.e., from CO 2 to CO, Au doping of the catalyst has the best photocatalytic effect on the process of catalytic CO 2 reduction to CO. The optimal reaction process catalyzed by these Cu-based nanoparticles is the direct transfer of the H atom of the H 2 O molecule to the O atom of the CO 2 molecule without surface bridging, which can determine the catalytic CO 2 reduction process, but the hydrogen evolution reaction cannot occur simultaneously. During the reduction of CO to CH 3 OH, we propose that the photocatalytic activity of these models still follows the order Cu 37 -Au > Cu 37 -Ag ≥ Cu 37 -Zn > Cu 37 -Pd > Cu 38 . Our research found that doping with Zn, Ag, and Au is beneficial to improve the activity of photocatalysts.