The growth of cellular Internet services has led to many resource-heavy applications, such as real-world games and virtual reality. Mobile Edge Computing (MEC) allows mobile devices to offload computational tasks to edge servers within the network, effectively alleviating the processing burden on mobile devices. Recent advancements in Unmanned Aerial Vehicle (UAV) and chip technologies have substantially increased UAVs’ computational capabilities, enabling the integration of UAVs with MEC systems, which are ideal for a wide range of use cases. This paper evaluates the performance of Reinforcement Learning (RL) algorithms, specifically Deep Q-Network (DQN) and State-Action-Reward-State-Action (SARSA), in the context of task offloading within dynamic MEC environments. In this scenario, each mobile device offloads multiple tasks to a UAV (edge server), aiming to optimize task allocation to minimize overall deployment costs, including energy consumption and delay. RL techniques are applied to approximate the optimal solution for these tasks. The implementation of this work can be found at GitHub .