The movement of rock landslides is often accompanied by disintegration, with extreme occurrences resulting in the transition from rigid motion to fluidized motion. This change complicates the prediction of rock landslide dynamics, as the mechanisms governing these two types of motion vary significantly. To address this issue, this study introduces a novel method for assessing the disintegration rate of rock masses using the principle of energy conservation. The rock disintegration terms are incorporated into two dynamic models, each describing the motion of the rigid and fluidized phases of a rock landslide, enabling the depiction of the transition from rigid to fluidized phases and its impact on landslide dynamics. A novel computing method based on the finite volume method is also proposed to solve the model equations and their coupling relationships. This method offers the advantage of solving the equations of two dynamic models using a unified grid system and a variable time step, reducing computational costs while maintaining precision. The back analysis of a laboratory experiment and the Tagharma rock landslide is conducted to validate the applicability and benefits of the proposed numerical approach. Alternative simulations suggest that the rock structure may determine the characteristics of rock landslide motion, leading either to rigid or fluidized motion tendencies. Moreover, the impact of rock disintegration on landslide motion, whether positive or negative, is contingent upon the extent of disintegration and the distance traveled by the fluidized phase.