Mobile Edge Computing (MEC) is considered a standard mechanism to minimize computing latency and task offloading by allowing mobile systems to offload their intensive operations. Non-orthogonal multiple access (NOMA) is an effective mechanism for improving spectrum efficiency. At the same time, the massive multiple-input multiple-output (MIMO) assists an enormous number of users with continuous offloading. Therefore, this work implements an intellectual task offloading process for MIMO-NOMA in MEC. This paper proposes a novel task offloading process for MIMO-NOMA in Mobile Edge Computing (MEC) that significantly improves energy consumption, delay, and system performance. By employing the Pine Cone Optimization Algorithm (PCOA), the proposed system reduces energy consumption by 34%, minimizes overall delay by 24.1%, and enhances achievable capacity by 30.8% compared to traditional models. The results demonstrate that the proposed system outperforms existing algorithms and models regarding bit error rate, total energy consumption, and average run time. The proposed task offloading process is particularly suitable for MEC applications that are latency-sensitive and energy-consuming. This work provides a new perspective on integrating MIMO-NOMA with MEC, enabling efficient task offloading and improving system performance.