In the context of inflation, a model is said to be predictive if the predictions depend on a number of parameters smaller than the number of predictions themselves. In this talk we consider different types of single field inflationary models (such as Higgs inflation-type and Cosmological Attractors), discussing the relationship between their unitarity and renormalizability properties and the quest for predictiveness. In order to consistently connect the low and high energy regimes of the effective field theories (described in terms of Standard model and inflationary parameters respectively), some threshold corrections are needed. This raises the question: how are the predictions sensitive to the required UV completion? We show under which circumstances the observables predicted by the model are (in)sensitive to the threshold corrections.