We survey the transport of proteins across nanopores in the framework of coarse-grained modeling. The advantage of a reduced complexity with respect to full-atomistic techniques lies in the possibility of massive sampling of events, thus allowing a statistical mechanical description of translocation in terms of ensemble averages. Often, protein transport through narrow channels tightly couples with unfolding pathways causing a richer phenomenology compared to unstructured polymer translocation. This reflects into a process controlled by the presence of protein-specific free-energy barriers which can be conveniently estimated by statistical mechanical methods implemented in coarse-grained simulations. We illustrate how protein transport dynamics can be characterized by the statistical properties of trajectories and sometimes interpreted as driven diffusion of a single collective coordinate over a free-energy landscape. We also discuss, through selected examples, the connection between reduced-model simulations and recent experimental results.