Free API key · Works on n8n Cloud

Free text to speech for n8n, Make and Zapier

Send text from any workflow and get an MP3 back, usually within a second or two. It runs on the HTTP Request node that comes with n8n, so there is nothing to install, and a free FreeTTS key covers 15,000 characters a month.

Download the n8n workflow Get a free key

Last updated 17 October 2026 by the FreeTTS team · Tested in n8n 2.43.7 · Every sample on this page was made with the API call it describes

Free key, no card
Nothing to install
318 free voices, 2,253 on PRO
MP3 link back in one call

FreeTTS news bulletin Ran in 5 s

Run twice a daySchedule Trigger · 0 6,18 * * *

Write the scriptEdit Fields · text, voice

FreeTTS: make the audioHTTP Request · POST /api/v1/tts

Download the MP3HTTP Request · Response: File

Outputbinary · data · audio/mpeg

0:25

"voice": "en-GB-SoniaNeural",
"chars_used": 392,
"audio_url": "https://freetts.org/api/audio/…"

Your FreeTTS key for n8n

n8n sends this key with every request, so FreeTTS knows whose characters to count. It is shown once, so paste it into n8n straight away.

Credential type: Header Auth
Name: x-api-key
Value: your key

n8n keeps credentials encrypted and leaves them out of downloaded workflows, so you can share the workflow file without sharing the key.

Hear the difference

The same headline, three voices. Press play.

“This is the six o'clock news. Here are tonight's headlines.”

FreeSoniaStandard voice · English (UK)

0:04

PROAdaHD voice · English (UK)

0:05

UltraDukeUltra voice · on freetts.org

0:03

A free key reads with the 318 standard voices and adds a short spoken FreeTTS tag at the end of each file. A PRO key adds the HD and Signature voices, with no tag. Ultra voices like Duke come with PRO on freetts.org, and are not on the API.

Quick facts

What it is: A REST endpoint that turns text into speech. Your workflow sends the text and a voice name, and FreeTTS answers with a link to the MP3.
Endpoint: POST https://freetts.org/api/v1/tts
The key: One header, x-api-key. In n8n that is a Header Auth credential.
What comes back: JSON with audio_url, the characters you have used this month and your limit, the voice that spoke, and whether a spoken tag was added.
The MP3: Waiting at audio_url for 30 days. Anyone with the link can download it.
Free key: 15,000 characters a month, 5,000 a day and 5,000 per request, 10 requests a minute, the 318 standard voices, MP3. Each file ends with a short spoken tag. For personal, non commercial use.
PRO: $19 a month: 1,000,000 characters a month, 10,000 per request, 200 requests a minute, HD and Signature voices, WAV and OGG, no tag, commercial use.
Works in: n8n Cloud and n8n on your own server, Make, Zapier (on a paid Zapier plan), cron, and anything else that can send an HTTP request.
Tested: The workflow file on this page was imported and run in n8n 2.43.7 on 17 October 2026.

n8n

Set it up in n8n: four nodes

Import the ready workflow, or build it by hand in about five minutes. Either way you end up with the same four nodes.

Fastest

Import the ready workflow

Open a new workflow in n8n, open the menu with the three dots at the top right, and choose Import from URL or Import from File. You can also copy the file's contents and paste them straight onto the canvas.

https://freetts.org/text-to-speech-for-n8n/freetts-news-bulletin.json Download the workflow file

Inside: both triggers, the script, the two HTTP Request nodes and a note with the three things to change. Then add your key and press Test workflow.

Or build it by hand

Schedule Trigger for when it runs
Edit Fields for the words and the voice
HTTP Request to FreeTTS, which makes the audio
HTTP Request that downloads the MP3

Add a Manual Trigger next to the schedule, so you can press Test workflow without waiting for six o'clock.

1
Schedule Trigger
When the workflow runs.
Trigger Interval
Custom (Cron)
Expression
0 6,18 * * *
That is 06:00 and 18:00 in the workflow's time zone, which you can change in the workflow settings. Hourly is 0 * * * *, weekdays at 07:30 is 30 7 * * 1-5.
2
Edit Fields
The words, and who reads them.
text
String · your script
voice
en-GB-SoniaNeural
In a real workflow the text usually comes from a node in front: RSS Read for headlines, Google Sheets for a list of lines, an AI node that writes the script, or an HTTP Request to a weather API.
3
HTTP Request: make the audio
The only call to FreeTTS.
Method
POST
URL
https://freetts.org/api/v1/tts
Authentication
Generic Credential Type, then Header Auth, then your FreeTTS credential
Send Body
On · Body Content Type JSON · Specify Body: Using JSON
JSONn8n expression
```
{
  "text": {{ JSON.stringify($json.text) }},
  "voice": "{{ $json.voice }}"
}
```
JSON.stringify keeps quote marks and line breaks in your script from breaking the request.
4
HTTP Request: download the MP3
Turns the link into a file.
Method
GET
URL
{{ $json.audio_url }}
Options
Add option, Response, then Response Format: File
Output field
data
The MP3 is now a binary file in data. Send it anywhere a node takes a file:
- Google Drive · Upload
- Dropbox · Upload
- FTP · Upload
- Telegram · Send Audio
- Gmail · attachment
- Amazon S3
- Read/Write Files from Disk

What FreeTTS sends backJSON

{
  "file_id": "d7f491c4-1924-4e86-b61a-08fb0ecc4df3",
  "format": "mp3",
  "audio_url": "https://freetts.org/api/audio/d7f491c4-1924-4e86-b61a-08fb0ecc4df3",
  "chars_used": 392,
  "chars_limit": 15000,
  "voice": "en-GB-SoniaNeural",
  "downgraded": false,
  "watermark": "end"
}

audio_url: The MP3. Download it in the next node. It stays available for 30 days.
file_id: The file's ID, if you prefer to build the link yourself.
chars_used: Characters used this month, this request included.
chars_limit: Your monthly allowance: 15,000 on a free key.
voice: The voice that actually spoke.
downgraded: True only if a Signature voice hit its monthly cap and a matching voice stood in.
watermark: "end" or "mid_end" on a free key, "none" on PRO.

Hear it

What comes back

The six o'clock bulletin three ways, all made with the API call above on 17 October 2026: first exactly as a free key returns it, tag included, then the same script in an HD voice and a Signature voice from a PRO key. Below it, two more everyday jobs in free voices.

The six o'clock bulletin, three waysThe same 392 characters in a free, an HD and a Signature voice

Good evening. This is the six o'clock bulletin. The city council has approved a new cycle lane along the river, with work starting in the spring. Train services return to normal tomorrow morning after weekend engineering works. And in sport, the town club stays top of the league after Saturday's win. Now the weather: light rain this evening, clearing overnight, with a low of eight degrees.

SoniaFree keyStandard voice · ends with “Generated with FreeTTS dot org”

0:28

AdaPRO · HDHD voice · English (UK) · no tag

0:28

CharonPRO · SignatureSignature voice · English (UK) · no tag

0:24

A free key file of a minute or longer carries the tag once in the middle as well. Start PRO free for 3 days and hear your own script in these voices.

Morning weatherAva · English (US) · Free voice

Good morning. Here's the weather for today. Sunny spells this morning, then cloud building after lunch, with a small chance of a shower around four. A high of nineteen and a low of eleven. If you're cycling home, the wind picks up after five.

0:16

A daily briefingAndrew · English (US) · Free voice

Here's your briefing for today. You have three meetings, and the first one is at half past nine with the design team. Two invoices are due this week. And the reminder you asked for: call the printer about the new menus before noon.

0:13

Make

The same thing in Make

Make (formerly Integromat) does it with its own HTTP modules. Five modules, no code.

Any trigger. For fixed times, open the scenario's schedule and pick Every day at a time. For news, start with RSS, Watch RSS feed items.
JSON, Create JSON. Build a small structure with two fields, text and voice. It escapes quote marks and line breaks for you.
HTTP, Make a request. URL https://freetts.org/api/v1/tts, method POST, a header x-api-key with your key, body type raw with content type JSON, and the JSON string from step 2 as the request content. Turn on Parse response.
HTTP, Get a file. URL: map audio_url from step 3.
Where it goes. Google Drive, Upload a file, or Dropbox, or FTP, with the file from step 4.

Zapier

And in Zapier

Zapier needs one less step, because file fields accept a link and Zapier fetches the file itself.

Trigger: Schedule by Zapier. Every Day, at the hour you want. Or any app trigger, such as a new RSS item or a new row in Google Sheets.
Action: Webhooks by Zapier, POST. URL https://freetts.org/api/v1/tts, Payload Type json, Data text with your script and voice with en-GB-SoniaNeural, Headers x-api-key with your key.
Action: Google Drive, Upload File. For File, map Audio Url from step 2. Dropbox, Slack and Gmail work the same way.

Worth knowing: Webhooks by Zapier is a premium app and this Zap has three steps, so it needs a paid Zapier plan or Zapier's free trial. Make's free plan and n8n on your own server can do it at no cost.

No platform

Cron, curl and a few lines of code

If a server or a Raspberry Pi is already running, you don't need an automation tool at all. Set FREETTS_KEY once and schedule the script.

bulletin.shsh

#!/bin/sh
# bulletin.sh: turn a script into bulletin.mp3 with FreeTTS
URL=$(curl -s https://freetts.org/api/v1/tts \
  -H "x-api-key: $FREETTS_KEY" \
  -H "Content-Type: application/json" \
  -d '{"text": "Good evening. Here is the evening bulletin.", "voice": "en-GB-SoniaNeural"}' \
  | python3 -c 'import sys, json; print(json.load(sys.stdin)["audio_url"])')
curl -s -o bulletin.mp3 "$URL"

crontabcron

# crontab -e   (runs at 06:00 and 18:00 server time)
0 6,18 * * * FREETTS_KEY=your_key /home/you/bulletin.sh

Times follow the server's clock. Check it with date before you rely on six o'clock.

bulletin.pyPython

import os
import requests

r = requests.post(
    "https://freetts.org/api/v1/tts",
    headers={"x-api-key": os.environ["FREETTS_KEY"]},
    json={"text": "Good evening. Here is the evening bulletin.",
          "voice": "en-GB-SoniaNeural"},
    timeout=60,
)
r.raise_for_status()
audio = requests.get(r.json()["audio_url"], timeout=60)
open("bulletin.mp3", "wb").write(audio.content)

JavaScript, Node.js, subtitles and the full reference are on the developer page.

Ideas

What people build with it

Real jobs FreeTTS keys do every day, with the number that decides whether the free plan is enough: characters a day.

A newsreader for an internet radio station

A small internet radio station in England reads its own news this way. At six in the morning and six in the evening a script gathers the latest headlines and the local weather, sends about 460 characters to FreeTTS, and the station plays the MP3 on air. Nobody sits at a microphone, and it has run twice a day, every day, since it was set up.

Job	Trigger	Words come from	Voice	Characters a day	Free key covers
Radio news bulletin	06:00 and 18:00	RSS Read for headlines, a weather API	Sonia	about 920	16 days a month
Morning weather announcement	Every day at 07:00	A weather API through HTTP Request	Ava	about 250	The whole month
Your own daily briefing	Weekdays at 07:30	Google Calendar, Gmail and an AI node	Andrew	about 600 on weekdays	The whole month
Smart home announcements	A Home Assistant or ioBroker event	The event itself	Jenny	about 150	The whole month
Voice replies in Telegram or Discord	A new message	An AI Agent node	Andrew	about 1,000	15 days a month
Voiceover for a Short or ReelPRO for paid use	A new row in Google Sheets	The sheet	Aria at +10%	about 900 a video	16 videos a month
Podcast intro and outroPRO for paid use	Every Monday	A template with the episode title	Ava HD	about 1,500 a week	Needs PRO for HD

Free audio is for personal, non commercial use. A monetized channel, a station that runs ads, or client work needs PRO.

Voices

Voices for news, weather and announcements

The same line in eight voices. Copy an ID straight into the voice field. Six are free; the last two need PRO.

SoniaEnglish (UK)

Free