Background and aim: ADMET prediction, 2D-, 3D-, and 4D-QSAR, QM/MM calculations, quantum chemical calculations, Molecular Dynamics, and Molecular Docking simulations have been covering a wide area in drug discovery. This theoretical approach enables to predict the mode of interaction of a ligand with its receptor. However, synthetic inhibitors were the most efficient ones; indeed, Gliptine is the most commercialized treatment of type 2 diabetes. On the other hand, natural inhibitors have also provided good activities via Dipeptidyl peptidase-4 (DPP-4) enzyme (p-Cymene, Thymoquinone, Carvacrol, α-Pinene, β-Pinene, Limonene). Method: In this work, we presented a theoretical investigation of dipeptidyl peptidase-4 enzyme inhibitors by natural inhibitors including the salvation parameter. To this, we used different molecular modeling approaches as molecular mechanics. Results: The molecular dynamics study was conducted for the best natural inhibitors obtained from the simulation of molecular docking with the lowest energy scores for ligands: Lig5, Lig3, and Lig ref. The interactions of our target and the studied inhibitors were further evaluated by using the molecular docking/dynamics simulations. A few key residues (GLU205, GLU206 and H-donor, ionic) were identified at the binding site of DDP-4. Conclusion: The obtained result of both methods, docking and dynamics molecular lead to the same conclusion and it was predicted that dithymoquinone (essential oil Nigella Sativa) presents better interaction of DDP-4 Enzyme in the presence of water in case of molecular dynamics and consequently can be the best inhibitor candidate to be investigated in vivo and in vitro.