Is there any more insight into what I'm paying for with the "VAPI Platform" here? Does any of this come from AI inference costs? If its truly majority of my expense is just coming from VAPI server hosting to glue together twilio + other api calls, I will be forced to move off the platform as my app scales up (building voice agent at consumer app scale) https://cdn.discordapp.com/attachments/1211484481848873020/1408983125710864515/Screenshot_2025-08-23_at_3.13.24_PM.png?ex=68abb94a&is=68aa67ca&hm=a2f30da2d42f304a787d70749ab76276757733949d8fa25248236c28c2ccf469&