ElevenLabs

ElevenLabs Conversational AI runs your agent (voice, persona, knowledge, tools) and exposes a WebSocket per session. The bridge proxies Teler call audio into that WebSocket and relays the agent’s response back.

Reference repos

Python teler-elevenlabs-bridge

Node teler-elevenlabs-node-bridge

Prerequisites

• An ElevenLabs account with a Conversational AI agent created
• 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
ELEVENLABS_WEBSOCKET_URL	yes	none	The agent’s WebSocket URL, e.g. wss://api.elevenlabs.io/v1/convai/conversation?agent_id=<id>
ELEVENLABS_SAMPLE_RATE	no	16k	Audio sample rate. 8k or 16k.
TELER_API_KEY	yes	none	Your Teler API key from the dashboard.
NGROK_AUTHTOKEN	yes (local)	none	ngrok auth token for the public tunnel.

Tip

The agent’s behavior (voice, tools, prompt, knowledge base) is configured in the ElevenLabs dashboard, not here. The bridge is just an audio pipe.

5-minute setup

Python

git clone https://github.com/frejun-tech/teler-elevenlabs-bridge
cd teler-elevenlabs-bridge
cp .env.example .env
# fill ELEVENLABS_WEBSOCKET_URL, TELER_API_KEY, NGROK_AUTHTOKEN
docker compose up --build

Node

git clone https://github.com/frejun-tech/teler-elevenlabs-node-bridge
cd teler-elevenlabs-node-bridge
cp .env.example .env
docker compose up --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"
  }'

Architecture

Same 3-endpoint shape as every Teler recipe. See Bridge for the conceptual model. The bridge holds two WebSockets: one to Teler, one to ElevenLabs, and pumps audio between them.

Production deployment

• Replace ngrok with a real public host (Fly.io, Railway, ECS, your own).
• Pin the Docker image rather than :latest.
• Move ELEVENLABS_WEBSOCKET_URL to your secrets manager (it embeds your agent_id).
• Forward webhooks to your backend rather than the bridge for durability.

Troubleshooting

Symptom	Likely cause	Fix
Agent answers but voice is wrong	Wrong agent_id in WebSocket URL	Verify the URL; agent must exist in your account
Phone rings then drops	Bridge unreachable from public internet	Check ngrok; verify SERVER_DOMAIN
Choppy audio	Sample rate mismatch	Try ELEVENLABS_SAMPLE_RATE=8k
Agent doesn’t recognize speech	Background noise, low input volume	Test with a clean audio source first

Next steps

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

Bridge architecture The conceptual model behind every recipe.

ElevenLabs Conversational AI Configure your agent on the ElevenLabs side.