Abstract: In this study, a new method based on a chemical thermodynamic model is proposed to quantitatively predict the phase distribution in cemented soil. Commercial kaolinite with simple phase composition is selected, and a series of cemented kaolinite samples are prepared by adjusting cement content, binder-water ratio, and NaOH/KOH doses. The samples are then characterized qualitatively and quantitatively using X-ray diffraction (XRD) and ²⁹Si-nuclear magnetic resonance (²⁹Si-NMR) techniques, respectively. Then, a thermodynamic model of cemented kaolinite is constructed using the GEMS software and the chemical thermodynamic database Cemdata18, simulating the evolution of mineral distribution in cemented kaolinite under different cement contents and alkali activation conditions. The simulation results are compared and analyzed with the micro-characterization results. The research indicates that the thermodynamic model can reliably predict the mineral distribution evolution in cemented kaolinite. Meanwhile, cement and alkalis have varying maginitude of "dissolution-polymerization" effects on the high polymerization chemical groups in soil, and the type of alkali significantly affects the depolymerization of clay minerals and the type and structural characteristics of the polymerization products.