XML Web services' distributed computing model allows applications to communicate with each other, regardless of where or how they are implemented. Thus, Web services-based distributed systems technology facilitates the development of distributed, service-oriented information retrieval system over the heterogeneous Internet. This paper presents a Web services- based model for information retrieval across different platforms. The network topology structure, service calling process, task allocating and load balancing as well as time-threshold setting are described. The characteristics of the model are briefly addressed.