In deep and ultra-deep complex formations, strong heterogeneity poses challenges to predictability. The acquisition, transmission, and integration of formation-engineering data are complex, influenced by intricate subsurface conditions. Additionally, the accuracy of physical modeling for drill ability of formations is constrained, and the uncertainty in rock-breaking mechanics further complicates matters. Traditional spatial interpolation methods struggle to ensure modeling accuracy, especially in cases of abrupt changes in formations. To address these challenges, this paper proposes an attribute-constrained method for modeling formation drill ability based on seismic and borehole data. Under the constraint of seismic attributes, the limited drill ability information from wells is interpolated and extrapolated according to the geological “facies” characteristics represented by attribute descriptions. A three-dimensional formation drill ability model is established, providing a better simulation of variations and uncertainties in deep formations.