This usually can be fixed by selecting a different model temporarily and then switching back to the one you want to use.. Give it a try and if it doesn't work then logging out and back in again should solve it..
#I'm getting a persistent "The selected model can not be used for text to speech" error
visual yew