Triggering TTS

REST API

The fastest way to speak a message from any external app or script.

POST http://localhost:39231/api/send?token=YOUR_TOKEN

{
  "type": "tts",
  "params": {
    "value": "Wow, this tutorial is just way too cool"
  }
}

With a specific voice and volume:

{
  "type": "tts",
  "params": {
    "value": "Hello from the API",
    "voice": "Brian",
    "volume": 75
  }
}

voice matches either the voice's id or label from the voices list. volume is 0–100 and applies on Windows only.

See Get a token if you don't have a token yet.

Chatbot commands

Add a TTS tab to any chatbot command in the Lumia UI. The message field supports the full variable system:

{{username}} just subscribed for {{subMonths}} months!

The voice and volume set on the command override the global defaults for that specific firing.

Alerts

Every alert has a TTS tab in the alert editor. The message resolves against the alert's event variables ({{username}}, {{amount}}, {{message}}, etc.), the same ones available in the alert's display text.

When sending an alert via the REST API you can override the TTS message in extraSettings:

{
  "type": "send-alert",
  "params": {
    "value": "twitch-subscriber",
    "extraSettings": {
      "username": "lumiastream",
      "tts": "lumiastream just subscribed!"
    }
  }
}

Custom JavaScript

From a Custom JS action, use lumia.tts() to speak a string directly:

await lumia.tts({
  message: `Welcome back, ${lumia.getVariable('username')}!`,
  voice: 'Brian',
  volume: 80,
});

The call resolves when the message is queued (not when it finishes playing). To wait for playback to finish before continuing the action chain, set waitForTts: true on the command's TTS tab in the UI.

Variable function

You can fire TTS inline in any template using the speak variable function:

{{speak=Hello from a variable function}}

This works anywhere variables are resolved: chatbot responses, alert text, overlay templates.

Plugin SDK

From a plugin, use this.lumia.tts():

await this.lumia.tts({
  message: 'ElevenLabs audio ready',
  voice: 'default',
  volume: 90,
});

For plugins that generate their own audio (ElevenLabs, TTS Monster, etc.), use this.lumia.playAudio() instead; those plugins intercept the normal TTS pipeline and produce audio bytes directly. See the ElevenLabs TTS example for the full pattern.

Muting TTS

Method	Effect
{{set_sfx_muted=true}}	Mutes TTS, HFX audio, and AudioManager globally. Use in a template, action, custom JS, or bind it to your own chatbot command.
{{set_sfx_muted=false}}	Re-enables the above.