Integrations
KugelAudio voices in Vapi
KugelAudio has a built-in Vapi custom TTS endpoint. Point your assistant at it with one voice configuration and the calls use a KugelAudio voice, with no server of your own in between.
How it works
What the integration does
No proxy server
Vapi calls the KugelAudio endpoint directly, once per phrase, and gets raw PCM16 back at the sample rate it asked for.
The key stays out of the URL
Put the API key in server.secret. Vapi sends it as the x-vapi-secret header on every request.
Voice chosen in the URL
The voice_id query parameter selects the voice. Add model_id to pick the model explicitly.
Code
The assistant voice configuration
Send this with PATCH /assistant/{id}, or include it when you create the assistant with POST /assistant. Replace the voice ID and the API key with your own.
{
"voice": {
"provider": "custom-voice",
"server": {
"url": "https://api.kugelaudio.com/vapi/synthesize?voice_id=YOUR_VOICE_ID",
"secret": "YOUR_KUGELAUDIO_API_KEY",
"timeoutSeconds": 30
}
}
}Copied from the documentation at docs.kugelaudio.com/integrations/vapi. Read the full guide
Integrations
Other integrations
LiveKit Agents
Official TTS plugin for LiveKit Agents in Python and TypeScript.
View integrationPipecat
Official TTS service for Pipecat voice pipelines in Python.
View integrationCognigy
Custom speech vendor endpoint for Cognigy Voice Gateway.
View integrationMCP server
Connect Claude, Cursor or any MCP client and generate speech from a chat.
View integration
Give your agent a KugelAudio voice
Create an account, copy an API key from the dashboard and pick a voice. Every integration on this page runs on the same API.