Deep coalbed methane (CBM) reservoirs are characterized by low porosity and low permeability. The complexity of hydraulic fracture networks controls reservoir stimulation efficiency, while the in-situ stress field controls fracture initiation and propagation. To address the challenges posed by the insensitivity of in-situ stress to elastic parameters and the limited accuracy of conventional seismic inversion methods, this study proposes an integrated workflow comprising “rock mechanical parameter inversion—heterogeneous modeling—stress field simulation.” Taking deep coal seams in the northeastern Ordos Basin as the study area, spatially variable mechanical parameters, including Young’s modulus, Poisson’s ratio, and density, were obtained through pre-stack seismic direct inversion. These parameters were subsequently input into a finite element model to simulate the three-dimensional in-situ stress field of the No. 8 coal seam of the Taiyuan Formation. The results show good agreement between the simulated values and measured data from six wells, with relative errors for the maximum and minimum horizontal principal stresses below 4.6% and 6.9%, respectively. The study reveals the spatial heterogeneity characteristics of the in-situ stress field in the No. 8 coal seam and identifies areas with a horizontal stress difference of 3–8 MPa as favorable “sweet spots” for hydraulic fracturing. This method overcomes the limitation of conventional simulations that rely on well interpolation for mechanical parameter assignment, which inadequately captures heterogeneity, and achieves an integrated framework combining seismic exploration and geomechanical analysis, providing a reliable scientific basis for fracturing optimization and sweet spot prediction in deep CBM reservoirs.