Borehole collapse pressure prediction plays a key role in drilling safety, reducing construction costs, and realizing efficient drilling. Fracture development under complex ultra-deep geological conditions significantly affects the prediction of borehole collapse pressure. Conventional methods rely on finite element simulations for 3D geomechanical modeling and 3D collapse stress prediction, which although, highly accurate, requires substantial computational resources. To address this issue, the study proposes an efficient and rapid in situ stress modeling method driven by seismic data, utilized for 3D collapse pressure prediction. Initially, a combined spring model with curvature properties is developed using a multi-scale data of pre-stack seismic and rock mechanics logging to model a three three-dimensional in situ stress field efficiently and rapidly. Next, based on the maximum likelihood attribute, the fracture development was obtained from 3D seismic data to provide 3D weak surface attribute parameters for the study area. Finally, the collapse model of sliding along the fracture plane was calculated using the Mohr-Coulomb criterion. This enables the collapse pressure prediction of the fractured formation from one-dimensional logging data to a three-dimensional working area. This method was applied to the woodworking area of Tari, with results showing a high agreement between model predictions measured data, reaching 93.79%. The prediction results also aligned well with formation micro-resistivity scanning imaging interpretations, verifying the method’s feasibility for predicting borehole wall collapse events. This study demonstrates that rapid, high precision modeling of collapse pressure can provide an integrated geological engineering solution for drilling in ultra-deep and complex areas.