Recent advances in genome-scale metabolic network reconstruction paved the way to the use of quantitative modelings such as FBA. However, despite the great interest of these techniques to tackle quantitative features, microbial community modeling remains unclear. Whereas studies represent a microbial community with several compartments for each microbial strains and their common pool, others advocate for the use of a single compartment that combines all reactions. Here we show that both modelings lead to different optimal quantitative solutions. This study illustrates this difference by the use of the flux module technique, that describes, in a compact way, the optimal solution space as computed by FBA-like techniques. For application, this paper computes the flux modules of a hot spring microbial community (represented by Synechococcus spp., Chloroflexus and Roseiflexus spp.) and a microbial methagenic system (Desulvovibrio vulgaris and Methanococcus maripaludis) sul-fate reducing bacteria), while emphasizing the quantitative changes that occurs when one assumes either the consortium as a " single compartment " or a multiple compartment.