Contact-based methods are employed used to monitor landslides, but their high cost, inefficiency, and reliance on discrete points limit a comprehensive understanding of landslide deformation behavior. To address this limitation, a geodetic inversion strategy integrating creep and dislocation models is proposed to estimate landslide creep and slip from InSAR observations, enabling the quantitative separation of their respective contributions to landslide motion. The method was applied to the Xiongba landslide in Gongjue County, Tibet. SBAS-InSAR-derived ascending and descending deformation of the Xiongba landslide was used to estimate creep and slip models via the proposed method. The results show that the Xiongba landslide is dominated by slope-aspect motion, with a maximum rate of 495 mm/yr, and secondary landslide bodies H1 and H2 exhibit distinct movements but share similar viscosity parameters. The inversion results indicate that surface deformation of the Xiongba landslide is predominantly controlled by slip processes, whereas the contribution of mass creep is comparatively limited. Model misfits of 7.9 (ascending) and 4.7 (descending) mm/yr indicate good agreement with observations, and the alignment between estimated and observed slip depths further supports the model’s reliability. The proposed inversion framework provides a new means of disentangling creep and slip contributions to landslide deformation and offers valuable insights into the failure mechanisms governing large-scale landslides.