Anyone has any idea on how to build a TTS and STT eval pipeline for VAPI?

To integrate your chosen STT, LLM, and TTS providers via vapi customizable API, then log timestamps and outputs at each stage to measure latency, accuracy, and quality, automate test calls using vapi API, configure endpointing/interruption settings, and track real-time performance metrics to continuously optimize speed, clarity, and conversational flow.. https://docs.vapi.ai/how-vapi-works https://docs.vapi.ai/customization/voice-pipeline-configuration

Thanks. But current TTS and STT are not perfoming very well.

@piyushgambhir you found a way ? looking for the best stt