In 1791, one year after the first US Patent Act, the French ‘Loi relative aux découvertes utiles’ sought to revolutionize the Ancien Régime’s privileges of invention, turning them into new patents or ‘brevets d’invention’. Not only did the law recast ci-devant privileges theoretically, declaring the new patents to derive from natural rights rather than royal grace, but it did so also in a more practical way. Whereas privileges of invention were only delivered after a long and tedious process of prior examination, mostly conducted by the royal administration and by members of the Paris Academy of Sciences, patents were delivered on request. The only condition was that would-be patentees file a description of their invention—through writing, drawing, or other modes of representation such as three-dimensional models or samples. Although the law insisted that descriptions be ‘exact’ and encompass ‘all the means one proposes to employ’, in the absence of any guidelines or formal requirements from the administration, inventors were largely left to fend for themselves when they attended to the description of their inventions. Drawing from established scholarly conventions of describing the arts and crafts, from past and present patent specifications circulating from England, and soon from the expertise of the first patent agents, these early patentees used the description of their inventions not only to meet (fuzzy) legal requirements, but also to further their business interests and claims for cultural status. In doing so, they helped to shape the meaning of ‘intellectual’ in IP. Through both a qualitative study of several early patents and a quantitative analysis of all the patents that were delivered between 1791 and 1844, when a new law was passed, this presentation recovers the diversity of media, styles and strategies that French inventors chose to adopt when they faced the necessity of, and opportunities offered by, describing their inventions. Specifically, I argue that in the case of France, drawing rapidly became the favored medium through which inventions were to be described; more precisely, analytic and schematic drawings formed a regime of description through which objects were not merely re-presented, but instead were abstracted into diagrams that generated objects. To explore how this regime of description shaped the appropriation of inventions and the understanding of the ‘intellectual’ in IP in France, I contrast this situation with the case of inventors and patents in the United States, where models, not drawings, were the favored medium for the description of inventions—eventually becoming a compulsory requirement in the 1836 Patent Act. While many actors, on both sides of the Atlantic, readily agreed that models, compared to drawings, were more material, and thus closer to the objects they represented, the consequences that this specific regime of description had for appropriation were far from straightforward; paradoxically, models fostered a kind of IP that was much more ‘intellectual’ than the one that drawings had helped generate in France.