Magnetorheological (MR) gels are a new class of MR composite materials that can effectively address the issue of particle agglomeration in MR fluids, enhancing the controllability of materials in engineering applications. In this study, a novel parametric model was developed to track the nonlinear hysteretic behaviours with strain-stiffening of the MR gel. The model parameters were identified using the measurement data pertaining to five current levels (0A, 0.2A, 0.5A, 0.8A, and 1A) under a strain amplitude of 10% and a frequency of 0.1 Hz. The fruit fly optimisation algorithm (FOA) was used to determine the optimal solution for the model parameters. The model parameters were generalised with respect to the loading current, and the reliability of the generalised model was verified. The results of the study demonstrate that the proposed model outperforms two classical models, the Bouc–Wen model and viscoelastic–plastic model, in capturing the nonlinear and strain-stiffening behaviour of MR gels, particularly for higher current excitations. This study has potential applications in predicting the nonlinear hysteresis behaviour of automotive dampers, and provides a theoretical basis for the semi-active control of suspensions.