Large gap of supply and demand for mineral resources brings an urgent need to develop technology and methods for mineral exploration. Metallogenic prediction plays a leading role in mineral exploration. In consideration of current software platform we used and characteristics of specific metallogenic prediction method, the efficiency and quality of metallogenic prediction are not so satisfactory. This article describes an open source spatial information software-based technical framework and how it is used to develop the system for metallogenic prediction. The architecture of the framework is clearly stated in this paper, based on the study of principles of GIS-based metallogenic prediction. An application on copper mineral prediction in the Yangtze River metallogenic belt is discussed, which is executed by using the proposed framework and one of the existing GIS software platform respectively. The results prove that our proposed framework has significantly improved the efficiency of metallogenic prediction, on premise of that the prediction result is accurate. It provides a free convenient way for metallogenic prediction.