🇬🇧 EN
🇮🇹 IT

3.7 ElevenLabs

Configure ElevenLabs text-to-speech voices for use in AI Voice Call questionnaire channels. ElevenLabs provides high-quality, natural-sounding voices — including the ability to clone custom voices — that replace the default Twilio synthesis when configured.

API Key Configuration

  1. Open ElevenLabs settings Navigate to Settings → ElevenLabs in the sidebar.
  2. Enter your API Key Paste the API key from your ElevenLabs account (elevenlabs.io → Profile → API Keys). The key is stored encrypted.
  3. Save Click Save. Once the key is saved, Flusso fetches your available voices from the ElevenLabs API.

Voice Cloning

ElevenLabs supports voice cloning — you can create a custom voice from audio samples directly in your ElevenLabs dashboard. Once a cloned voice is available in your ElevenLabs account, it will appear alongside the built-in voices in Flusso after saving your API key.

Cloned voices: Voice cloning is managed entirely within the ElevenLabs platform. Flusso reads all voices available on your account (both built-in and cloned) and presents them for selection.

Voice Selection for Questionnaire Voice Calls

Once your ElevenLabs API key is configured, ElevenLabs voices become available in the questionnaire Voice Call channel settings. When configuring a questionnaire for AI Voice Call delivery:

  1. Open the questionnaire editor Navigate to Questionnaires and open the questionnaire you want to configure.
  2. Open Voice Call settings Click the Voice Call button in the editor top bar to open the voice configuration drawer.
  3. Select an ElevenLabs voice In the voice selection dropdown, choose any of your ElevenLabs voices. The dropdown lists both built-in and cloned voices from your account.
  4. Save Save the questionnaire. Outbound voice calls for this questionnaire will now use the selected ElevenLabs voice for text-to-speech synthesis.
Tip: Test your chosen voice with a short questionnaire before deploying to production. Different voices have varying characteristics for clarity, warmth, and pacing that may suit different use cases.