This paper investigates the Heterogeneous Electric Vehicle Routing Problem with Time Windows and Energy Minimization, a critical challenge in sustainable urban logistics. We propose a Mixed-Integer Linear Programming model that explicitly minimizes the total energy consumption of a heterogeneous electric vehicle (EV) fleet while satisfying customer demands within specific time windows. The model accounts for vehicle diversity in terms of battery capacities, consumption rates, and charging characteristics, and incorporates distance- and load-based energy consumption under a full-recharge policy. Computational experiments are conducted on a comprehensive set of benchmark-based instances. Results demonstrate the model’s ability to find optimal or near-optimal solutions for small-scale problems, with acceptable optimality gaps for larger instances. Sensitivity analyses reveal the influence of key parameters—battery capacity ratios, consumption rates, recharging speeds, maximum number of active EVs, and payload sensitivity—on energy usage and fleet configuration. Notably, increasing battery capacity or fleet size significantly reduces energy consumption up to a saturation point, while inefficient or slow-charging vehicles negatively affect performance. The findings have direct implications for logistics operators seeking to enhance energy efficiency and environmental performance in EV-based delivery systems.