For the computational researches on the large-scale metabolisms, MetaGen, a user-friendly utility to batch process and model the organism-specific multi-level metabolisms in KEGG into enzyme and pathway graphs, was developed. An example was given by expanding a little on the bow-tie structure of metabolic networks to show MetaGen is a useful and promising tool. MetaGen takes advantages of KEGG web service to ensure data reliability, uses relational database to accelerate the modeling process and facilitate data management, and applies advanced software modeling techniques as well as a pluggable software architecture for the easy expansion of functionalities in the future. MetaGen saves researchers from the elaboration on preparing metabolic networks for computation, which paves the way to deepen the researches on the large-scale metabolic networks. MetaGen is fully open-sourced and available at http://bnct.sourceforge.net/.