Hi guys, if your voice agent is linked to to Make.com and being used for cold calls, is there a way to make it more snappy in replies. Also, to cut out inconsistencies - sometimes it pauses for a full 10 seconds or more.

What does your set up look like? Do you have a knowledgebase? Tools? What LLM & audio models are you using? 10 seconds means something is definetly wrong lol

Use streaming mode with persistent connections, enable audio caching, choose low-latency ASR/TTS like AssemblyAI, use smaller/faster LLMs with short prompts, enable Vapi’s endpointing and interruption features..

@Kings_big💫 Streaming mode?

Streaming mode delivers real-time, partial responses over a persistent connection, enabling fast, natural voice interactions with minimal delay.. https://docs.vapi.ai/chat/streaming

She is not using the chat API.

The streaming mode sends real-time, partial responses via SSE or WebSocket, enabling fast, natural voice or chat interactions with low latency..

@Shubham Bajaj is this guy just sending AI responses

Thank you very much for the replies guys - very helpful 😊

Sorry, your question looks confusing..

Yeah, unfortunately

@elsa231 @SlaviSavanovic Apologies for the delay. Hey! To help track down this issue, could you share: - The call ID - When exactly this happened (the timestamp) - What response you expected to get - What response you actually got instead This would really help us figure out what went wrong!