There is evidence suggesting that condensation trails (contrails) are at least as warming as carbon dioxide for aviation. Contrail warming effects can be reduced with contrail avoidance strategies. A key element necessary to develop effective contrail avoidance techniques are contrail models. Such models require speed to be implemented into dispatcher workflow, and a minimum accuracy to ensure a net environmental benefit. No models have been proven to be suitable yet. To understand the capabilities of current models, two existing models of different fidelities, CoCiP and APCEMM, are compared in this investigation under certain and uncertain inputs. The comparison of the models under uncertainty is necessary to determine the model sensitivity at different timescales. Further, propagating uncertainties is required for their primary use case since weather data are highly uncertain. The findings presented reveal that the models display similar behaviors when varying the relative humidity and ambient temperature, but they otherwise disagree in the evolution of the contrail properties with time. An increased amount of appropriate validation data is necessary to create a model suitable for real-life contrail avoidance. This investigation also estimated the minimum accuracy required for a contrail avoidance model to be 65 %.