Efficient and accurate numerical simulations of cooled high-pressure turbines are crucial for improving thermal load predictions and optimizing turbine designs. The choice of gas model significantly influences the accuracy of these simulations, with different models requiring varying levels of numerical effort. This study investigates the impact of different gas models on the prediction of thermal loads, and consequently the cooling design of a high-pressure turbine stage derived from an ultra high bypass-ratio geared turbofan engine concept. Using DLR’s turbomachinery CFD solver TRACE, three simulations with varying gas models were conducted to evaluate their influence on the wall temperature prediction. The turbine stage, optimized in a dedicated design process, features a realistic film-cooling configuration applied to both the nozzle guide vanes and rotor blades. A two-dimensional temperature distribution was imposed at the inlet to represent the hot-streak emerging from the combustor. The results demonstrate that temperature-dependent gas models have a significant impact on the predicted thermal loads, highlighting their importance for accurate cooling design. Conversely, the inclusion of pressure-dependent gas models was found to have negligible influence on the thermal loads for this specific turbine design. These findings underscore the importance of selecting appropriate gas models to ensure accurate predictions while optimizing computational efficiency. The study concludes that temperature-dependent gas models are essential for reliable thermal load predictions in cooled high-pressure turbines, while pressure-dependent models may not be necessary for efficient and accurate turbine cooling designs that improve engine performance and durability.