Hello, I am currently using ElevenLabs with my own credentials on VAPI and I have two questions: 1. Are you using TTS service or Conversational service when we use ElevenLabs for TTS in VAPI? 2. Do you know if there is a way to track cost related to each call on ElevenLabs when using VAPI with our own credentials? Regards,

Hi there! 👋 I can answer both of your questions about ElevenLabs integration with VAPI: 1. VAPI uses the Text-to-Speech (TTS) service from ElevenLabs for its TTS functionalities. This integration allows you to leverage ElevenLabs' speech synthesis capabilities, including voice model selection and parameter tuning for natural-sounding voice applications. [Learn more about this integration here](https://docs.vapi.ai/providers/voice/elevenlabs). 2. Yes, there are several ways to track costs related to each ElevenLabs call when using VAPI with your own credentials: * **End-of-Call Webhooks**: VAPI provides webhooks that send detailed reports after each call, including cost information. [Check the webhook documentation here](https://docs.vapi.ai/server-url/events#end-of-call-report). * **Analytics API**: You can use VAPI's Analytics API to systematically track usage and group call data by assistants. [See the Analytics API documentation](https://api.vapi.ai/api#/Analytics/AnalyticsController_query). * **VAPI Dashboard**: View and manage call costs directly through your VAPI dashboard at [dashboard.vapi.ai/calls](https://dashboard.vapi.ai/calls). For reference, the approximate cost breakdown is around $0.04/min for ElevenLabs voice plus VAPI's charge of $0.05/min, totaling about $0.09/min. Let me know if you need any clarification or have additional questions! Sources: - [VAPI ElevenLabs Integration](https://docs.vapi.ai/providers/voice/elevenlabs) - [End-of-call-report documentation](https://docs.vapi.ai/server-url/events#end-of-call-report) - [Analytics API](https://api.vapi.ai/api#/Analytics/AnalyticsController_query) - [VAPI Dashboard](https://dashboard.vapi.ai/calls)

> 1. Which ElevenLabs service is being used in VAPI for TTS? VAPI is using ElevenLabs' Text-to-Speech (TTS) service, not the Conversational service. > 2. Tracking cost related to each ElevenLabs call when using VAPI with your own credentials Yes, VAPI does track costs for ElevenLabs calls, even when using your own credentials. Here's how it works: 1\. **Cost Tracking Implementation**: - VAPI tracks the number of characters processed by ElevenLabs TTS - It has a

numCharactersUsed

property that increments as text is sent to ElevenLabs 2\. **Integration with VAPI's broader cost tracking**: - Costs from ElevenLabs are collected through the

api

function - VAPI adjusts costs when using your own credentials (BYOK - Bring Your Own Key) - These costs are logged and tracked for analytics purposes 3\. **Viewing Cost Data**: - The costs are logged and can be viewed in call analytics - They appear in the call-level statistics under provider costs However, it's important to note that while VAPI tracks these costs internally, you will still be billed directly by ElevenLabs for the usage through your own API key. VAPI's cost tracking is for informational purposes to help you understand the costs associated with each call. The cost tracking in VAPI helps you associate specific calls with their ElevenLabs usage, but the actual billing relationship is between you and ElevenLabs since you're using your own credentials. Is there anything specific about the cost tracking implementation you'd like me to elaborate on?