Metformin is the best therapeutic choice for treating type 2 Diabetes.Despite this, and the fact it has been prescribed worldwide for decades, its mechanism of gluconeogenesis inhibition is still unknown.In the following work a novel mechanism of inhibition is suggested: that metformin performs its action on the target enzyme not as a pure molecule but, after sequestering endogenous cellular copper, as a copper complex.This result was obtained using chemoinformatics methods including homology modeling for the creation of the target enzyme’s tridimensional virtual structure, molecular docking for both the determination of the movement of the prosthetic group inside its cavity and for the identification of the best ligand poses for the metformin copper complexes, and eventually pharmacophore modeling and virtual screening to find alternative virtual leads that could achieve similar effects.The simulations show the complex binding as a non competitive inhibitor to the large exit of the mitochondrial glycerophosphate dehydrogenase enzyme’s FAD cavity, preventing FAD movement inside the cavity and/or quinone interaction and therefore its electron transfer function.The proposed mechanism seems to be successful at explaining a wide range of existing experimental results, both regarding measurements of metformin non-competitive inhibition of GPD2 and the role of copper and pH in its action.The virtual screening outcome of at least two similarly active purchasable molecule hints to an easy way to experimentally test the proposed mechanisms.In fact, the virtual leads are very similar to the copper complex but quite different from metformin alone, and a laboratory confirmation of their activity should plausibly imply that metformin acts in synergy with copper, giving us the ability to design new antidiabetic drugs in a novel and more rational fashion, with significant savings in research costs and efforts.