The growth of renewable energy requires reliable storage solutions, and vanadium redox flow batteries (VRFB) meet this need. The diffusion of vanadium ions through membranes influences the capacity of VRFB. When diffusion rates differ between half-cells, this causes electrolyte imbalance and promotes self-discharge, resulting in a loss of storable energy. The 3D-printed SPEEK membrane (3D-SPEEK) modifies this diffusion compared to conventional membranes, offering different performances. To analyze these effects, a dynamic model was used, based on the molar balance of the four vanadium ions and the Nernst equation. The model was implemented in Python with NumPy for numerical calculations. Diffusion coefficients were experimentally measured for four membranes: Nafion211, Fap450, conventional SPEEK, and 3D-SPEEK, allowing for a comparison of their effects on performance. Simulations show that the capacity loss over cycles directly depends on the specific diffusion coefficients of each membrane. The 3D-SPEEK membrane, with the lowest diffusion coefficients, shows the best capacity retention. Its energy efficiency remains stable over the long term and the variation of ionic concentrations is regular. These results indicate that this membrane is promising for VRFB and show that the model accurately captures the influence of ionic diffusion on performance.