This study presents a sophisticated thermoeconomic model, meticulously integrated with Pareto optimization for the marine nuclear propulsion system, with the Nuclear Ship Savannah serving as a reference benchmark. Given the constraints in the available design data for the Savannah marine nuclear propulsion (SMNP) system, innovative ideas are devised to precisely characterize its thermodynamic properties, ensuring a rigorous and reliable analysis. Following model validation, multi-objective particle swarm optimization is applied to the SMNP to optimize performance criteria, such as energy efficiency and total product exergy cost rate , to identify Pareto-optimal solutions and evaluate performance trade-offs. The optimized Savannah marine nuclear propulsion (OSMNP) is determined through the analytic network process within a multi-criteria decision-making framework. A thorough comparative analysis of the SMNP and OSMNP is then performed from the energy, exergy, and exergoeconomic perspectives, incorporating key system indicators, including energy efficiency and total product exergy cost rate , exergy efficiency , total propulsion power , and total capital cost rate . The results reveal that SMNP achieves and of 26.18% and 51.8%, respectively, with corresponding and of $249.07/h and $4616.8/h. In addition, reaches 15.09 MW. In contrast, OSMNP exhibits a remarkable improvement in , , and but a marginal rise in and . The and increase to 27.82% and 54.51%, representing enhancements of 1.64% and 2.71%, respectively. The also experiences a significant boost, reaching 15.93 MW—an increase of 0.84 MW. These performance gains are achieved with only a small rise in to $253.8/h (a minor increment of $4.73/h) and to $4990/h (an increase of $373.2/h). Moreover, the OSMNP demonstrates a significant reduction in the relative exergy cost difference of components, indicating a notable decrease in exergy destruction and inefficiencies compared to the SMNP. Finally, to assess the effectiveness of particle swarm optimization (PSO) in solving the SMNP problem, a comparative analysis is conducted against the firefly algorithm and genetic algorithm using the Wilcoxon rank-sum test. The comparative analysis demonstrates that PSO outperforms the other optimization algorithms. These findings support the development of sustainable, environmentally friendly marine propulsion technologies.