The model answers with what answer returns for the request, for an answer that depends on what was asked.