Since XML technology was launched on the world in 1998, there has been an increase in the number of XML documents generated from business transactions. Consequently, both industry and academia have proposed several new kinds of database technologies (i.e. Native XML database (NXD) systems) as an efficient approach for storing, and processing XML documents. This paper presents a comparative analysis of the two storage models namely text-based and model-based used by Open Source Native XML databases to manage their internal storage mechanisms. Additionally, it specifically examines performance evaluation of Berkeley DB XML, Xindice, eXist, and dbXML from the point of view of handling database query operations. Each product is deployed and evaluated using some benchmark techniques. A number of interesting insights are gained from these experiments and a summary of the strengths and weaknesses of each of these products is presented.