Two different alternatives over conventional leaf spring are developed in this work, such as composite and hybrid composite-based leaf springs. In composite type, carbon-epoxy based multi-leaf springs are used and the combination of metal and carbon fiber material is used to develop hybrid composite leaf springs. The outputs namely deflection, maximum stress, and fatigue tests are determined using the finite element analysis approach. Besides, the optimal outcomes are set as the target for hybrid deep neural network-rat swarm optimization based prediction. The findings shows that the hybrid composite leaf spring provides better and enhanced results than the composite leaf spring. © 2021 Taylor & Francis Group, LLC.