Abstract: The formation drillability grade is an important parameter in drilling engineering, which is reflected in mud logging, well logging and seismics data. Accurately predicting the drillability grade is of great significance for optimizing drilling parameters and improving drilling efficiency. Against this background, this paper focuses on the drilling operations of deep and ultra-deep wells in Block BZ19-6. This study adopts the formation drillability grade as the key evaluation indicator and aims at the enhancement of drilling speed and efficiency. A multi-source information database covering pre-drill seismic, mud logging while drilling, and post-drill well logging is established. By using the hierarchical-range composite analysis method, dominant information indicators of drillability are screened out. Subsequently, PSO-BP neural network algorithm is engaged to construct an intelligent prediction model for well profiles through fusing mud logging and well logging data. Finally, the spatial distribution prediction of drillability is realized combining with the model-based post-stack seismic inversion method and composite seismogram time-depth conversion calibration. The findings show that the dominant factors of drillability in the target area from well logging data are acoustic transit time, rock density, neutron porosity, gamma ray, and resistivity, while the dominant factors from mud logging data are weight on bit, rate of penetration, drilling fluid viscosity, displacement, d_c index, and relative torque. The root mean square error(0.385 2) based on dominant factors of well logging is remarkably lower than that(0.444 3) of mud logging prediction. There obviously exists a response relationship between formation drillability and seismic attributes such as amplitude and frequency. The drillability grade values in the target area inverted based on the fusion of well and seismic data mainly range from 2.81 to 8.61. The lithologies with high drillability grade are mainly conglomerate, tuff, and interbedded tuffaceous sedimentary rock. This method achieves full-cycle fine evaluation of formation drillability grade, providing a basis for the optimization of drilling parameters and enhancement of drilling efficiency in deep complex formations.