It is relatively easy for us to provide a (potentially long) list of scientific models. But can we say (or define) what a scientific model is? In my talk, I’m going to argue that a descriptive account of scientific models (e.g. one of distinguishing different types of scientific models) is not going to help us with that question. I will then discuss and ultimately dismiss two philosophical (or normative) accounts of scientific models: one that answers the above question by considering the relation between models and theories, and one that defines scientific models in terms of structures. I will finally put up for discussion whether we can make progress when paying attention to the functions that scientific models serve.