Abstract We compare the outcome of the Copenhagen and the Berlin statistical multifragmentation models. The two models both start from the assumption of an ensemble of various excited fragments at thermal and chemical equilibrium for an expanded nuclear system at a certain stage, the freeze-out configuration. Details as the volume of this source, the intrinsic excitation of the fragments, and the secondary decays are treated quite differently. Also the technical level of the Monte Carlo sampling differs. Main characteristics such as total charged multiplicity, IMF multiplicity, and total neutron yield are in reasonable agreement. Thus the overall predictions are quite robust against details in the description. Deviations are however seen for high energies ( E★ > 6 A MeV) where fewer IMF's, fewer neutrons, and much more α's are seen in the Copenhagen model. The isotopic distribution of the very light fragments is more narrow than in the Berlin model.