With the emergence of numerous new applications and the explosive growth of Internet-of-Things (IoT) devices in beyond 5G (B5G) networks, the massive yet delay-sensitive personalized content delivery has imposed a crucial challenge to mobile network operators (MNOs). Cooperation among MNOs for sharing the communication, caching, and computing (3C) three-dimensional resources in an economic yet real-time manner has provided a promising solution to address this challenge. In this paper, we investigate the 3C resource sharing among multiple MNOs to realize efficiently and economically personalized content delivery, where a third-party 3C resource provider (CRP) is introduced to manage the sharing 3C resource pool. By leveraging the multi-dimensional contract theory, we propose an optimal 3C resource contract scheme for the CRP in a realistic asymmetric information scenario, and the appointed 3C resources in one contract will be allocated to the MNO who signs it. For each MNO, we establish a partial transcoding model to achieve the optimal orchestration on the 3C resources, where the closed-form solution of caching placement and transcoding strategy is obtained. Then, MNOs can choose the most suitable contracts to sign based on the service requirements from end users. In particular, we analyze the global incentive compatibility and feasibility of the proposed multi-dimensional contract approach, which is theoretically proven to achieve the optimal solution. Extensive simulation results demonstrate the efficiency of the proposed 3C resource sharing mechanism compared with other benchmark schemes. Specifically, the proposed 3C resource sharing scheme can reduce 35% delivery delay compared with the non-sharing scheme.