A process for creating a media sequence, for example video or audio, and an associated device includes the stages of: obtaining a selection of media files; regrouping the chosen media files into a plurality of groups according to at least one coherence criterion; selecting, for each group of media files assembled in this way, a media motif among a plurality of predefined media motifs, the chosen media motif being able to define a media rendering for the set of media files of the group; and assembling the media motifs selected in this way and their associated media files in such a way as to form at least one definition of the media sequence. The process can relate to automated video editing as implemented on Web servers.