In this paper, the authors developed a model predictive control (MPC) algorithm for multi-port modular multilevel converters (MP-MMCs). MP-MMCs are used to enable the use of hybrid energy storage devices (HESDs) in a scalable energy management system (EMS) for electric vehicle (EV) applications. HESDs refer to the use of multiple types of energy storage in the same EV drivetrain system. In this paper, battery cells are sized for the EV energy density while ultra-capacitor cells are used for high acceleration periods. This system will reduce the size of the EV drivetrain due to the elimination of high-power inverters and its filtering components. Using MPC, this system can achieve the following control objectives: 1) extend the battery cells lifetime and driving range by shielding them from high power pulses, 2) balance the state of charge (SOC) levels of every storage cell, 3) increase the system efficiency through optimizing the supplied motor voltage and reducing the switching losses. Moreover, the proposed solution provides means for onboard high-power charging of EV storage cells. Validation results are provided in the paper using a developed hardware prototype along with co-simulations and hardware in the loop system to verify the system's effectiveness.