The durability of rocks is a substantial rock property that has to be considered for designing geotechnical structures. Uniaxial compressive strength (UCS) and Young's modulus (E) are key indexes for measuring rocks’ durability. Several types of artificial intelligence (AI) methods have been used for modeling these key indexes; however, surprisingly, no explainable AI (XAI) has been considered for their model developments. An XAI is a model whose assessment is not a black box, and humans could understand its problem solution approach. This study has filled this gap and presented SHAP (Shapley Additive Explanations) as one of the most recent XAI methods for modeling UCS, and E. SHAP value could successfully illustrate intercorrelations between rock properties (porosity, point load index, P-wave velocity, and Schmidt hammer rebound number) and their representative UCS and E for each individual record and also together as variables. Results indicated that P-wave velocity has the highest importance for UCS and E prediction. eXtreme gradient boosting (XGBoost) was used as a solid predictive AI system for UCS and E estimation. Outcomes (R²> 0.99) confirmed the high accuracy of the SHAP-XGBoost model comparing with other typical AI models (Random Forest and Support Vector Regression). These results indicated XAI could be considered for illustrating complicated relationships within rock mechanics and energy-resource developments.