Is it possible to get Vapi interaction into voice-to-text-response mode and save on time by switching off TTS stage? What if I want to make iOS app which sends speech via SDK, but shows only text responses? Can I turn off TTS or point it to some dummy endpoint? Any ideas?