Vapi

Vapi runs your end-to-end voice agent: STT, LLM, TTS, conversation logic, tool calls. The bridge proxies Teler call audio into Vapi and relays the agent’s response back.

Reference repos

Python teler-vapi-bridge

Node teler-vapi-node-bridge

Prerequisites

• A Vapi account with an assistant created and deployed
• The assistant_id of that Vapi assistant
• A Teler account with an active API key
• One Teler phone number on your account
• An ngrok auth token (for local development)
• Docker and Docker Compose

Configuration

Variable	Required	Default	Description
VAPI_API_KEY	yes	none	Your Vapi API key.
VAPI_ASSISTANT_ID	yes	none	The ID of the Vapi assistant to connect calls to.
VAPI_SAMPLE_RATE	no	16000	Vapi-side sample rate in Hz.
VAPI_MESSAGE_BUFFER_SIZE	no	50	Queue depth before relaying.
TELER_API_KEY	yes	none	Your Teler API key.
NGROK_AUTHTOKEN	yes (local)	none	ngrok auth token.

Tip

The assistant (voice, prompt, tools, knowledge) is configured on Vapi’s side. The bridge is just an audio proxy.

5-minute setup

Python

git clone https://github.com/frejun-tech/teler-vapi-bridge
cd teler-vapi-bridge
cp .env.example .env
docker compose up -d --build

Node

git clone https://github.com/frejun-tech/teler-vapi-node-bridge
cd teler-vapi-node-bridge
cp .env.example .env
docker compose up -d --build

Initiate a call

curl -X POST http://localhost:8000/api/v1/calls/initiate-call \
  -H "Content-Type: application/json" \
  -d '{
    "from_number": "+91XXXXXXXXXX",
    "to_number":   "+91XXXXXXXXXX"
  }'

Audio quirks

Vapi defaults to 16 kHz. The bridge converts to Teler’s 8 kHz automatically. Keep VAPI_SAMPLE_RATE=16000 unless you have a specific reason to change it.

Tuning the buffer

VAPI_MESSAGE_BUFFER_SIZE controls how many in-flight messages can queue before relaying. Increase if you see message drops; decrease if latency matters more than throughput.

Troubleshooting

Symptom	Likely cause	Fix
Silent call	Wrong VAPI_ASSISTANT_ID	Verify in the Vapi dashboard
401 from Vapi	Invalid VAPI_API_KEY	Rotate the key in Vapi
Choppy / dropped audio	Buffer too small	Raise VAPI_MESSAGE_BUFFER_SIZE
Wrong voice / language	Assistant misconfigured	Edit the assistant in Vapi

Next steps

Other recipes OpenAI Realtime, ElevenLabs, Gemini, and more.

Bridge architecture The conceptual model behind every recipe.

Vapi docs Configure your assistant on the Vapi side.