Phase-field modeling has become a popular method for numerical simulation of fracture, thanks to its ability to capture complex fractures without any algorithms for tracking fracture geometry. Yet most of the existing phase-field models are not fully appropriate for fractures in geomaterials as they ignore several critical features including frictional contact, pressure-dependence, quasi-brittleness, and their combined impact on mixed-mode fracture. This talk will introduce novel phase-field models tailored to fractures in geomaterials, in which the aforementioned features are incorporated in a well-verified and validated manner. In particular, the double-phase-field model can simulate, and naturally distinguish between, cohesive tensile fracture and frictional shear fracture in geomaterials, without any additional algorithms for contact constraints as well as those for surface tracking. Applications of the phase-field models to various geomechanical problems will also be discussed briefly.