The Agent Voice Response (AVR) platform supports integration with ElevenLabs Speech-to-Speech (STS), enabling high-quality, real-time conversational AI. ElevenLabs is well known for its natural and expressive voices, making it an excellent choice for creating human-like conversational agents.
To configure the ElevenLabs STS service, set the following variables:
Variable Description Example Value
PORT Port on which the ElevenLabs STS service runs 6035
ELEVENLABS_AGENT_ID Your ElevenLabs Agent ID (required) your_agent_id
ELEVENLABS_API_KEY API Key (only required for private agents) sk-xxxx
Before using this integration, configure your ElevenLabs Agent with the following audio settings:
These settings are mandatory to ensure proper audio compatibility and real-time streaming performance with AVR.
avr-sts-elevenlabs:
image: agentvoiceresponse/avr-sts-elevenlabs
platform: linux/x86_64
container_name: avr-sts-elevenlabs
restart: always
environment:
- PORT=6035
- ELEVENLABS_AGENT_ID=$ELEVENLABS_AGENT_ID
- ELEVENLABS_API_KEY=$ELEVENLABS_API_KEY
networks:
- avr
avr-core:
image: agentvoiceresponse/avr-core
platform: linux/x86_64
container_name: avr-core
restart: always
environment:
- PORT=5001
- STS_URL=ws://avr-sts-elevenlabs:6035
ports:
- 5001:5001
networks:
- avr
You can find the official repository here:
Github: agentvoiceresponse/avr-sts-elevenlabs
For a complete example, check the docker-compose-elevenlabs.yml in the avr-infra project.