Compliant walking model served as a theoretical framework because of the generation of kinetic information of human walking. Moreover, as Whittington applied a curvy foot to the compliant walking model, the exaggerated ratio of the vertical ground reaction force (GRF) to the horizontal GRF and the fixed center of pressure during the single stance phase have been solved. However, giving meaning to the stiffness of the compliant leg became difficult because of the exaggerated height of the foot and the shortened leg length. Moreover, joint torque was not compared with the ankle torque because the joint position is different from the ankle position. Thus, in this study, we modified the compliant walking model with a curvy foot that reflects the configuration of the human leg and examined whether the modified compliant walking model with a curvy foot could represent human walking. We reflected the configuration of the human leg to the model by shortening the height of the curvy foot, lengthening the compliant leg length, and shifting the joint position similar to the ankle position. To confirm the feasibility of the modified model and to examine the effect of shifting the joint position, we compared the simulation data with the experimental data: the GRF obtained with the force plate and the ankle torque obtained with the inverse dynamics through kinematic data. Thus, the modified model can reproduce the symmetric M-shape of the GRF but not the asymmetric shape of the model. Furthermore, the modified model can also produce the joint torque profile similar to the ankle torque profile. In conclusion, considering the human leg configuration can provide us with information about human walking.