<p>This paper explores the generativity of generative design algorithms (GDAs). Generative design (GD) is a process in which designers assign some of their tasks to a computational tool to generate a set of design solutions. While GDAs have been heavily studied, few studies have focused on assessing their generativity; that is, their capacity to help designers create novel proposals that go beyond their initial knowledge. To address this gap, this research compares two GDAs, namely, NSGA-II and MAP-Elites, in terms of their capacity to generate Pareto fronts composed of highly varied design solutions (Pareto fronts with this property are called “splitting Pareto fronts” in this paper). Both algorithms are applied to the industrial design problem of constructing a battery layout for an electric vehicle. A statistical and empirical analysis of the design solutions generated is conducted. The results show that the Pareto fronts generated by MAP-Elites offer designers more degrees of freedom than those generated by NSGA-II do. Thus, the study highlights that the degrees of freedom afforded by GDAs depend on the working principles of the algorithms. From a practical point of view, the results of this study indicate that a GDA can artificially reduce the degrees of freedom of designers. This pressing issue is discussed to help designers make the best use of GDAs.</p>