<p>In control system design a considerable and satisfying approach is model order reduction, which endeavours to make straightforward lower order systems yet retain critical dynamic characteristics. Application of Genetic Algorithm to optimize the numerator variables and Salp Swarm Algorithm to optimize the denominator variables of the approximated model, is the proposal in this paper as a new hybrid soft computing technique for Model Order Reduction. In order to guarantee efficacy and stability, the proposed approach recommended to minimize the performance index between the response times for each step of the original and approximated models. To find the best numerator structure, Genetic Algorithm is used for its global search capacity, while Salp Swarm Algorithm helps via efficiently adjusting the denominator parameters using its adaptive leader follower dynamics. The importance and appropriateness of the suggested hybrid Genetic Algorithm-Salp Swarm Algorithm approach to generate high-fidelity reduced-order models with low Integral square error, superior time-domain response matching, and maintained system stability are demonstrated thru simulation results on benchmark high order systems. This method provides an appropriate replacement for simplifying models in intricate control systems.</p>