Cities are having a hard time untangling the complex transport system of today’s nation, which is a difficult issue having traffic jams, inefficiencies, and environmental problems that, although the number of people and transportation with them in metropolitan areas continues to rise. Besides that, the constraints of electric vehicles like range limitations or the availability of charging infrastructure are too much for traditional routing systems to handle at once. In addition to the waste of fuels and delivery delays that the inefficiencies are causing, these inefficiencies significantly increase the operating expenses and have a more negative impact on the environment as well. We propose a Deep Reinforcement Learning (DRL)-based intelligent routing architecture that integrates real-time, multi-modal data such as traffic patterns, weather, and charging station locations to resolve such issues. Besides that, this technology adapts to the ever-changing urban landscape as well as ensures that EVs are following the most energy-efficient routes while balancing battery consumption and trip time. Since the system is constantly learning from real-time data, it can make better, more sustainable decisions locally at any time. The DRL model is significantly better than routing methods such as Google Maps as it beats the latter in 53.11% of the cases, according to the test performed in 1000 simulation runs. At best, it achieved energy savings of up to 2.0661 units, with an average reduction of 0.4739 units per episode. The total energy savings of 213.2665 units by the system show the degree to which it is successful in upping EV route efficiency and showcasing the emergence of DRL in upgrading the urban transportation for a better and environment - friendly framework.