Cell-penetrating peptides enter cells via diverse mechanisms, such as endocytosis, active transport, or direct translocation. For the design of oral-delivered cell-penetrating peptides, it is crucial to know the contribution of these different mechanisms. In particular, the ability of a peptide to translocate through a lipid bilayer remains a key parameter for the delivery of cargos. However, existing approaches used to assess translocation often provide discrepant results, probably because they have different sensitivities to the distinct translocation mechanisms. Here, we focus on the passive permeation of a range of hydrophobic cyclic peptides inspired by somatostatin, a somatotropin release-inhibiting factor. By using droplet interface bilayers (DIB), we assess the passive membrane permeability of these peptides and study the impact of the peptide cyclization and backbone methylation on translocation rates. Cyclization systematically improved the permeability of the tested peptides while methylation did not. By studying the interaction of the peptides with the DIB interfaces, we found membrane insertion and peptide intrinsic diffusion to be two independent factors of permeability. Compared to the industrial gold standard Caco-2 and PAMPA models, DIBs provide intermediate membrane permeability values, closer to Caco-2. Even for conditions where Caco-2 and PAMPA are discrepant, the DIB approach also gives results closer to Caco-2. Thereupon, DIBs represent a robust alternative to the PAMPA approach for predicting the permeability of peptides, even if the latter present extremely small structural differences.