n8n text-to-speech in one HTTP node.

If your n8n workflow is already calling ElevenLabs and eating your credit balance, here's the swap. Drop-in HTTP node, download the workflow JSON below, ship it in about ninety seconds.

Start free → Get the workflow

The three-step swap.

Add an HTTP Request node in your workflow. Set method to POST, URL to https://api.voicepep.com/v1/audio/speech.
Set the body type to JSON. Paste the payload below with your Authorization: Bearer API key header. The node returns binary MP3 bytes.
Wire the output into your storage or upload node — Google Drive, S3, Dropbox, Slack, wherever your workflow already lands audio.

The exact HTTP node config:

{
  "method": "POST",
  "url": "https://api.voicepep.com/v1/audio/speech",
  "authentication": "genericCredentialType",
  "genericAuthType": "httpHeaderAuth",
  "sendHeaders": true,
  "headerParameters": {
    "parameters": [
      { "name": "Authorization", "value": "Bearer vp_live_your_key" },
      { "name": "Content-Type", "value": "application/json" }
    ]
  },
  "sendBody": true,
  "bodyContentType": "raw",
  "bodyContent": {
    "model": "standard",
    "voice": "am_michael",
    "input": "={{ $json.text }}"
  },
  "responseFormat": "file"
}

    The n8n gotcha: in the HTTP node, set Response Format to File (not JSON) — the endpoint returns raw audio bytes, not a JSON blob. If you leave it as JSON, n8n throws a parse error. That's the one thing every tutorial gets wrong.
  

Ready-to-import workflow.

Download this JSON, import it into n8n (Workflows → three-dot menu → Import from File), swap your API key, done:

Download workflow JSON →

Includes: a manual trigger, an HTTP Request node pre-configured for VoicePep, and a Write Binary File node that saves the audio to your n8n data folder. Ready in about ninety seconds.

The math, if you're currently on ElevenLabs.

Say your n8n workflow generates 400,000 characters of speech a month — a mid-sized podcast auto-narration or a faceless-YouTube script pipeline:

ElevenLabs Pro

$99/mo

VoicePep Creator

$23/mo, same workflow

Same voice quality (twelve warm voices, or upgrade to studio rendering on Agency+). Same n8n HTTP node. One URL and one key swapped in your existing workflow.

Common questions.

Do I need to change my whole workflow?

No — only the URL, API key, and (if you were using ElevenLabs) the payload field names swap. Everything downstream (storage, upload, publish) stays exactly the same. The endpoint returns MP3 bytes just like ElevenLabs.

What voices are available?

Twelve voices out of the box on every paid tier: Michael, Sarah, Lewis, George, Onyx, Bella, Heart, Nova, Alloy, Adam, Puck, Emma. Full list at /docs. Custom voice creation from Agency ($63/mo).

Will this work on n8n self-hosted?

Yes — n8n's HTTP Request node is identical whether you're on cloud, self-hosted, or Docker. The endpoint accepts requests from anywhere.

What if I hit my monthly character limit?

Top-up packs are $5 for 500,000 characters, $15 for 2M, $40 for 6M. Or upgrade the plan. Never a metered surprise; overage requires a click, not a threshold breach.

What next.

Grab a free API key on the pricing page (5,000 characters/month, no credit card), download the workflow JSON above, paste your key, run it. If it works — and it should, in under ninety seconds — upgrade to whichever tier matches your monthly volume.

Start free Full API docs →