The prediction of tip leakage flow aerothermal loss plays a crucial role in turbine preliminary design, which strongly affects the turbine performance. A new tip leakage loss model for unshrouded turbine is developed in this paper, considering the compressible flow aerothermal process inside the clearance. By coupling with a parameterized loading distribution model proposed in this work, in which the blade loading can be described by four independent parameters, Zweifel coefficient (Zw), diffusion factor (DF), peak velocity location (PVL), and leading edge acceleration (LEA), the loss model can evaluate the tip leakage loss in aerodynamic design process conveniently and accurately. The proposed models are validated by numerical simulations and the results show that, compared with the other acknowledged loss models, the loss model can reduce the deviations by more than 50%. Based on the models, the effects of blade loading design parameters on tip leakage losses are discussed via analysis of variance (ANOVA). The results show that Zw has the most significant influence. As Zw decreases from 1.0 to 0.7, tip leakage losses can be reduced by about 16%. Under lower Zw conditions, the joint effect of PVL and LEA is remarkable. However, under higher Zw conditions, the joint effect of the DF and PVL is more important.