The development of numerical models to analyze pathologies and implants related to the first metatarsophalangeal joint of the foot remains an issue for attention. The structural effects of implants pre-flexion have been discarded in several finite elements analyses due to complexities to achieve these positions. This work aims to evaluate if the pre-flexion stress state should be included or could be discarded when only flexion is applied in two different silicone commercial implants, Swanson and Tornier, during a gait cycle. Finite element models were created for silicone implants. Both models were discretized using high-order finite elements. The hyperelasticity constitutive material model of Arruda-Boyce was used, based on experimental data; its behavior was compared with linear elastic models reported and used frequently assuming small and large deformations and applying to the Swanson and Tornier implants a flexion angle of 64°, which corresponds to in vivo measurements reported after implantation. Comparison between models, regarding hyperelastic model, showed mean variations of up to 32.5% for stresses and 14.01% for bending moment in Swanson implant, while for Tornier implant mean variations of 29.73% and 632.55% was obtained for stress and bending moment respectively. The maximum stress value obtained for the hyperelastic model in the Swanson implant reached a value of 22.82% of the tensile strength of the implant material while in the Tornier implant reached a value of 25.92%, the above values were evaluated at a flexion angle of 64°. The results suggest considering in finite element analyses not only the stress state generated to achieve critical flexion position in pleflexed implants models but also the hyperelastic material behavior of silicone for implants to avoid dismissing the non-linear structural behavior of hyperelastic materials.