Peer-to-peer (P2P) networks consist of nodes where the communication and data sharing is carried on directly between nodes, rather than being arbitrated by an intermediary node. Each node in such a network can act as both a client and a server. The computing power from such a network can help solve computationally complex problems that require powerful supercomputers. This research proposes P2PCompute - a viable commercial model in the P2P computing field. It harnesses existing technologies including P2P, Java, and the universal description, discovery, and integration (UDDI) registry to enable distributed processing of tasks on multiple servers. It is well-suited to the heterogeneous environment on the Internet and can lead to the development of more commercial P2P computing applications. Using this model, clients can divide their tasks into sub-tasks and request servers on the Internet to run these sub-tasks.