Freight movement is an essential component of urban transportation system. However, regional freight data are costly and difficult to obtain. Though abundant freight data sources become available for online access and also turn out to be reliable for transportation research, they have been rarely utilized directly to freight transportation studies. The modeling results are seldom delivered to users in a timely and convenient way. This paper proposed a framework for Web-based freight data collection and modeling. It discussed feasible mechanics to incorporate online freight data sources into modeling practices. It also demonstrated the methods to present modeling results through web services, which can facilitate decision-making procedures in freight transportation planning. As a large metropolitan area with sophisticated transportation systems, the five-county Los Angeles was selected as a study area to illustrate the powerful functions of the web-based freight model. The empirical study showed that it is plausible to establish a web-based freight modeling system to integrated heterogeneous freight data and make modeling results accessible any time and any where through Internet.