The design of low-cost and high-activity oxygen reduction catalysts on the cathode is important to improve the conversion efficiency in proton exchange membrane fuel cells. A series of nonprecious 3d-TM-combined M 1 M 2 N 6 /C candidates are constructed within high-throughput screening to find out highefficient oxygen reduction reaction (ORR) catalysts through axially coordinating with 13 kinds of potentially accompanied ligands by first-principles calculation. Among the fabricated 55 pristine M 1 M 2 N 6 structures, some pre-3d metals (d e ≤ 5) of Sc-, Ti-, and V-contained M 1 M 2 N 6 /C are excluded due to poor ORR activities with too low negative limiting potentials (U L ) values (U L < −1.0 V). The screening results of three common ligands (R = −OH, −F, and −NH 2 ) coordination indicate that axial ligand modification could significantly enlarge the U L of M 1 M 2 N 6 −R. A volcano relationship between U L and ΔG of OH* intermediate is helpful to select four excellent ORR candidates of NiNi-, CoNi-, FeCo-, and CoCo-combinations to further perform 10 kinds of axial ligand engineering. More applicable ligands of −CO, −COH, −NH, −NO, -Cl, and -CH 2 are identified to enhance ORR activity for certain pristine M 1 M 2 N 6 catalysts. Totally, 16 different candidates exhibit good ORR activity with U L > 0.80 V of Pt(111) by ligand modification screening. The increased moderate positive charge on central metals is ascribed to the weakened interaction between catalysts and OH* in *OH-intermediates. Our work provides valuable insights into understanding ligand engineering on dual-metal catalysts, which is helpful for designing highly efficient, nonprecious single-atom catalysts by tuning the metal bonding environment.