Why is ElevenLabs text-to-speech API returning an empty audio response?

Below is a checklist of things to verify when you receive empty audio output even though the text field isn’t empty:

1. Verify the Request URL and Voice ID

Make sure you are calling the correct endpoint with the correct voice ID. The API expects the voice ID as part of the URL (i.e.

https://api.elevenlabs.io/v1/text-to-speech/{voice_id}

If the voice ID is missing or incorrect, the service might not process the request as expected.

2. Check the Text Field Expression

Your JSON uses extra quotes around the expression:

JSONCopied
"text":"'{{$(\'IF-ElevenLabs\').item.json.output?.replaceAll(\'"\',\"\"").replaceAll(\'\n\',\'\'\').replaceAll(\'\*\',\'\'\').replaceAll(\'-\',\'\'\').replaceAll(\':00\',\' horas \') }}'"

Unless those extra quotes are required by your n8n configuration, they may be causing the API to receive an unexpected string value. When the API receives text with unwanted leading or trailing quotes, it might not generate the audio as anticipated. Try removing the surplus single quotes so that your expression is injected directly into the JSON. For example:

JSONCopied
{
  "text": "{{$(\"IF-ElevenLabs\").item.json.output?.replaceAll('\"','').replaceAll('\\n','').replaceAll('*','').replaceAll('-','').replaceAll(':00',' horas ')}}",
  "model_id": "eleven_multilingual_v2",
  "voice_settings": {
    "stability": 0.5,
    "similarity_boost": 0.8,
    "style": 0.0,
    "use_speaker_boost": true
  }
}

3. Test with Static Text

Before relying on the expression, try replacing the text value with a static string to determine if the audio is generated correctly. For example:

JSONCopied
{
  "text": "Hello, this is a test.",
  "model_id": "eleven_multilingual_v2",
  "voice_settings": {
    "stability": 0.5,
    "similarity_boost": 0.8,
    "style": 0.0,
    "use_speaker_boost": true
  }
}

If the static text produces valid audio, then you know the issue lies within the expression or its formatting.

4. Review Voice Settings and Other Parameters

Ensure that the parameters in voice_settings (stability, similarity_boost, etc.) are within the acceptable ranges as described in the ElevenLabs documentation. An out-of-range value might cause processing issues, though the API would generally return an error rather than an empty audio response.

---

Additional Information Request

To help further troubleshoot the empty output—especially if the above steps do not resolve the issue—please provide details on your n8n environment if applicable:

• n8n version:
• Database (default: Supabase, Redis):
• n8n EXECUTIONS_PROCESS setting (default: own, main):
• Running n8n via (Digital Ocean):
• Operating system: Windows

These details might reveal if there’s any extra configuration or environment-specific behavior affecting the HTTP request execution.

---

By verifying the endpoint, cleaning up the expression (removing extra quotes), and testing with static text, you should be able to pinpoint why the audio response is empty. Let me know if you need further assistance!