Edge computing and artificial intelligence models in wireless communication networks represent a technological advancement, as edge computing brings resources, computing and processing capabilities, as well as services, closer to users. This combination improves network performance in terms of latency, bandwidth, and energy consumption. This paper focuses on the Deep Q-Network (DQN) algorithm, a deep reinforcement learning (DRL) approach applied in a multi-user and multi-server environment for task offloading to these servers. The goal of this approach is to determine the best offloading strategies in order to optimize energy consumption. Our approach reduces energy consumption by more than half compared to traditional offloading techniques, where all requests are either executed on the device itself or sent to a cloud computing center.