The paper deals with an evaluation of the multi-mechanism (MM) approach capabilities in the prediction of the cyclic behavior of different classes of metallic materials. For this objective, the tests detailed in (Taleb, Int J Plast 43:1–19, 2013a) have been simulated here by the MM model. In these tests, six alloys were considered: two ferritic steels (35NCD16 and XC18), two austenitic stainless steels (304L and 316L), one “extruded” aluminum alloy (2017A) and one copper-zinc alloy (CuZn27). The specimens have been subjected to proportional and non-proportional stress as well as the combination of stress and strain control at room temperature. The identification of the material parameters has been carried out using exclusively strain controlled experiments under proportional and non-proportional loading paths performed in the present study for each material. The model may describe a large number of phenomena with twenty five parameters in total but, it appears that for a given material under the adopted conditions, the activation of all parameters may be not necessary. Our attention was focused mainly on the capabilities to predict correctly the cyclic accumulation of the inelastic strain including the shape of the hysteresis loops. The comparison between test responses and their predictions by the MM model are generally satisfactory with relatively small number of material parameters (between eight and thirteen according to the material). One can also highlight the capability of the MM model to describe a transient ratcheting without activation of the dynamic recovery term in the kinematic variables. Finally, the MM model deserves improvement for a better description of the cyclic behavior of anisotropic materials.