.

To add backchanneling to your custom LLM in Vapi, use the configuration setting

"backchannelingEnabled": true

in your API request. For fillers, you can modify your system prompt to instruct the LLM to include conversational fillers like "umm" and "ahh" in its responses:

json
{
  "role": "system",
  "content": "Respond conversationally, using natural fillers like 'umm', 'ahh', and 'you know' occasionally to sound more human."
}

According to [Vapi documentation](https://docs.vapi.ai/how-vapi-works), this approach helps make LLM outputs sound more conversational. Sources: - [How Vapi Works](https://docs.vapi.ai/how-vapi-works) - [Custom LLM Tool Calling Integration](https://docs.vapi.ai/documentation/advanced-concepts/customization/custom-llm/custom-llm-tool-calling-integration) - [Discord Channel](https://discord.com/channels/1211482211119796234/1340031852819251384)

Could you please add a description to your support ticket.