<p>To address the pressing issue of CO₂ and NO emissions from marine vessels, this study aimed to develop high-performance MOF-based adsorbents tailored for ship exhaust treatment. A database of over 7,000 MOFs was constructed, enabling systematic analysis of the relationship between structural features and gas adsorption performance. Based on pore structure descriptors, machine learning models were established to predict the adsorption capacities of CO₂ and NO. Guided by the observed influence of metal centers, a new bimetallic Zn/Cu-MOF-74 was developed via a green solvothermal strategy. Zn/Cu-MOF-74 samples with different Zn:Cu molar ratios were synthesized and evaluated, confirming the 2:1 ratio as the optimal candidate. The Zn/Cu-MOF-74 (2:1) exhibited a CO₂ uptake of 2.82&#xa0;mmol·g ⁻¹ (5.57% higher than pristine Zn-MOF-74, 2.67&#xa0;mmol·g⁻¹) and a NO uptake of 0.70&#xa0;mmol·g⁻¹ (18.26% higher than pristine Zn-MOF-74, 0.59&#xa0;mmol·g⁻¹). The BET surface area was measured to be 400&#xa0;m²·g⁻¹, and the IAST selectivity for CO₂/NO reached 11.13 at 0.1 bar with an average value above 2.5 in the 0-1 bar range. IAST and breakthrough experiments further validated its superior CO₂ uptake and selectivity under realistic CO₂:NO = 85:15 conditions, surpassing the China Classification Society benchmark of 2.70&#xa0;mmol·g⁻¹ working capacity. This work highlights the synergy between data-driven design and environmentally friendly synthesis, with Zn/Cu-MOF-74 (2:1) meeting the VLCC-based working capacity benchmark of 2.70&#xa0;mmol·g⁻¹ and underscoring its potential for practical deployment in ship exhaust control.</p> Graphical Abstract <p></p>