This is because none of them have been deployed now. I think some models can be used via Inference Provider.