Skip to main content

Voice (TTS)

Open Sub Accounts → select a location → Voice (TTS) tab.

WhatGHL can turn outbound text into a spoken WhatsApp audio (MP3) using text-to-speech (TTS). The most common setup uses ElevenLabs, but the app may support other providers depending on your organization’s configuration.

Where to configure​

  1. Open the Voice (TTS) tab on the sub-account.
  2. Enable TTS for that location.
  3. Choose the provider and map voices per language (and regional variants if your UI offers them, e.g. Latin American country codes when configured).
  4. Complete any API keys or provider setup the screen requires. A missing configuration state usually blocks audio generation.

ElevenLabs setup​

Configure the ElevenLabs API key, pick voices per language (including regional variants), and test a sample message with the voice preview.

Sending audio with #voice​

In the GHL outbound message body, add #voice so the platform synthesizes speech instead of sending plain text. You can use #voice:xx to pick a language or voice route when your sub-account defines those keys.

Requirements in typical setups:

  • TTS enabled for the sub-account.
  • No conflicting interactive payload on the same send (polls, buttons, lists, etc.) — TTS messages are plain audio.
  • Reachable temporary media URL for WhatsApp to fetch the MP3 (your deployment should expose the API/media path expected by the product).

Full rules: Message codes from GoHighLevel (#voice / #voice:xx).

#voice:xx and #traslate:xx in GHL​

Send an outbound from GoHighLevel combining #voice:xx (synthesis voice and language) with #traslate:xx (translation to the target language) and see the result in WhatsApp.

Good practices​

  • Test one short phrase before using TTS on live customer threads.
  • Match voice and language to your brand; avoid switching voices randomly mid-journey unless intended.
  • Remember #voice is evaluated on the original composer text before translation strips markers — see the message-codes doc for trigger order.