ElevenLabs is the default text-to-speech provider in TeleQuick Voice. When you build a cascaded agent (speech-to-text → language model → text-to-speech) and don’t override the voice stage, the runtime synthesizes replies with ElevenLabs. You can also use your ElevenLabs key for streaming speech-to-text and for a full Conversational AI agent — all three plug in through the same per-tenant credential. This page shows the config. It is shipped and in production.
New to the runtime? Start with BYO ASR / LLM / TTS for the full cascade config file — this page is the ElevenLabs-specific slice of it.

​
Use ElevenLabs for text-to-speech

This is the common case: keep your speech-to-text and language model, and let ElevenLabs speak the replies.
1

Set the TTS stage in agent-config.yml

Set tts_provider: elevenlabs and put your ElevenLabs voice_id in tts_voice. Because ElevenLabs is the default, you can even omit tts_provider — but naming it keeps the config explicit.
agent-config.yml
name: support-agent
pipeline_mode: cascaded

asr_provider: deepgram
asr_model: nova-2

llm_provider: openai
llm_model: gpt-4o-mini

# --- Text-to-speech: ElevenLabs (default) ---
tts_provider: elevenlabs
tts_voice: <elevenlabs-voice-id>   # your ElevenLabs voice_id

language: en
system_prompt: >
  You are a friendly support agent for Acme. Keep replies short and spoken-friendly.
2

Add your ElevenLabs key

Provider keys are per-tenant. Set ELEVENLABS_API_KEY in the TeleQuick console at agent.telequick.dev — it is sealed at rest and injected at call setup, never baked into the config file. See Credentials.
3

Save and place a test call

Config is hydrated from the control plane with no rebuild. Place a call and check the latency breakdown — the TTS row shows time-to-first-audio for the ElevenLabs stage.

​
How it streams

The ElevenLabs TTS session opens a streaming connection (/stream-input), so audio begins flowing as the language model’s text arrives rather than after the whole sentence is generated — that is what keeps first-audio low. The default model is a low-latency turbo model; the synthesizer requests linear PCM, which the runtime resamples into its 8 kHz PCM16 call bus for you.
If a streaming session can’t open — a missing voice or a bad key — the runtime falls back to a one-shot HTTP synthesis for that turn so the caller still hears a reply instead of silence. Confirm the voice id and key up front so you stay on the streaming path.
tts_voice is your ElevenLabs voice_id (the long id from your ElevenLabs voice library), not the display name. Leave it empty to fall back to the provider’s default voice.

​
Use ElevenLabs for speech-to-text

The same ElevenLabs key also powers a streaming speech-to-text stage. Set asr_provider: elevenlabs to transcribe with ElevenLabs instead of the default:
agent-config.yml
asr_provider: elevenlabs

tts_provider: elevenlabs
tts_voice: <elevenlabs-voice-id>
Deepgram is the recommended default transcriber for English telephony (see Deepgram). Reach for ElevenLabs ASR when you want a single vendor and key across both the listening and speaking stages.

​
Use ElevenLabs Conversational AI

ElevenLabs Conversational AI (ConvAI) is a full duplex agent — ElevenLabs owns transcription, the language model, and the voice on one bidirectional connection, and the runtime bridges the call audio to it. Configure it as a single REALTIME node, the same way as the other speech-to-speech providers:
agent-config.yml
entry_node: REALTIME
pipeline:
  REALTIME:
    node_type: REALTIME
    provider: elevenlabs
    model: <elevenlabs-agent-id>   # required: your ElevenLabs Conversational AI agent id
    voice: <voice-id>              # optional: override the agent's configured voice
    instructions: |               # optional: override the agent's system prompt
      You are a friendly support agent for Acme. Keep replies short.
​
model
string
required
Your ElevenLabs Conversational AI agent id. This is mandatory — ConvAI dials a specific pre-built ElevenLabs agent, not a bare model name.
​
voice
string
Overrides the voice configured on the ElevenLabs agent. Leave empty to use the agent’s own voice.
​
instructions
string
Overrides the agent’s system prompt for this call. Leave empty to use the prompt set in ElevenLabs.
Public vs. private agents. A public ElevenLabs agent connects without an API key. A private agent needs ELEVENLABS_API_KEY in the tenant’s credentials.
ConvAI’s in-call tool / function calling is more limited than OpenAI Realtime. If your agent needs to call HTTP or MCP tools mid-conversation, prefer a cascade (where the LLM stage runs tool calling) or OpenAI Realtime.

​
Credentials

One key covers all three surfaces (TTS, ASR, ConvAI). Set it in the TeleQuick console at agent.telequick.dev; it is resolved from the control plane at call setup and sealed at rest. For single-tenant or local development, the runtime also reads the key from the environment as a fallback: 
export ELEVENLABS_API_KEY=...
An agent whose TTS (or ASR) stage has no usable ElevenLabs key is refused at session creation with a clean error rather than starting and failing mid-call. Confirm the key is present before you route live traffic.

​
Next steps

Full cascade config

Wire ElevenLabs TTS alongside your chosen ASR and LLM in one config file.

Speech-to-speech

Run ConvAI or another duplex model as a single REALTIME node.

Turn detection

Tune barge-in and backchannel handling around the ElevenLabs voice.

Latency breakdown

See TTS time-to-first-audio on a live call.