A key component of Bayesian optimization is the underlying probabilistic model used for the objective function. In practice, it is often unclear how one should appropriately choose this model, especially when gathering data is expensive. We introduce a novel automated Bayesian optimization approach that dynamically selects promising models for explaining the observed data using Bayesian optimization in model space. Crucially, we account for the uncertainty in the choice of model; our method is capable of using multiple models to represent its current belief about the objective and subsequently using this information for decision making. We argue, and demonstrate empirically, that our approach automatically finds suitable models for the objective function, which ultimately results in more-efficient optimization.