MacNorris Logo
ENElevenlabs Icon
Tool

ElevenLabs: AI Voices and Your Own Voice via API

ElevenLabs is an AI platform for speech. You put in text and get spoken audio, in a voice from the library or in your own. We used it to build the voice for our comic generator. Here is what it does, what it costs and where it hurts.

What is ElevenLabs?

Text in, voice out. Your own too.

ElevenLabs is an AI platform for speech. The core is text-to-speech: you write a text, ElevenLabs reads it aloud in a natural voice.

Further tools come on top, such as voice cloning, voice design, dubbing and speech-to-text. They run through the same account and the same credits.

With voice cloning you train a voice of your own. We read out a text that ElevenLabs displayed, and after about three minutes the material was enough. That voice now speaks in our comic videos.

3minutes of reading aloud were enough for us to get a voice of our own.
10,000credits per month on the free plan. Voice cloning is not included there.
2,200credits used in episode 1, including all tests.
What is ElevenLabs good for?

Four use cases ElevenLabs covers.

Four use cases ElevenLabs covers.

Voiceover for videos

You write the script, ElevenLabs speaks it. For explainer videos, demos and social clips this replaces the recording.

The result: Voiceover without a microphone and without a studio.

Your own voice as a tool

With voice cloning your voice becomes a building block. Record once, and ElevenLabs voices any text in your voice afterwards. Only clone voices you have the rights to.

The result: Your voice in every video, even when you are not recording.

Dubbing into other languages

With dubbing a video can be carried over into other languages while the voice stays recognisable. We have not tested this yet.

The result: One video for several markets.

Automation via API

ElevenLabs has an API for text-to-speech. You send text and get audio back, in any voice that lives in your account.

The result: Voiceover from scripts and workflows.

ElevenLabs costs

Free to Business: who needs what?

Free: 10,000 credits per month, without a commercial licence and without voice cloning. Good for testing, not for publishing.

Starter (approx. 6 USD per month): 30,000 credits, commercial licence and Instant Voice Cloning. The starting point if you want to publish videos.

Creator (approx. 22 USD per month): Over 100,000 credits and Professional Voice Cloning, for a voice in higher quality.

Pro, Scale and Business (approx. 99, 299 and 990 USD per month): Significantly more credits, from Scale on several seats and more voice clones. Enterprise by arrangement.

For us: In episode 1 we used around 2,200 credits, including all tests. The audio for a 10-second video cost about 0.10 EUR.

As of October 2026. Prices and credit amounts change, so check the current pricing page. What is often overlooked: Anyone who publishes needs at least a paid plan, because the free plan has no commercial licence.

Where ElevenLabs is strong

What ElevenLabs genuinely does well.

Getting to your own voice is quick. For us, around three minutes of reading aloud were enough, and then the voice was there.

Everything runs under one account: text-to-speech, voice cloning, dubbing and speech-to-text, with the same credits.

The API fits into workflows. Text in, audio out, which you can call from FastAPI or n8n.

Where ElevenLabs hits its limits

What ElevenLabs is not. And where it hurts.

Credits are hard to estimate. Usage depends on model and text length, and tests count too. We used around 2,200 credits in episode 1. Plan some reserve.

Publishing needs a paid plan. The free plan has no commercial licence and no voice cloning.

The microphone must be allowed. While training our voice we read away diligently without realising the browser had no access to the microphone. Check the permission before you record.

A cloned voice is a responsibility. Only clone voices you have the rights to. Your own is the safe case.

ElevenLabs and automation

From the voice into the workflow.

In our comic generator Notion provides the text, FastAPI sends it to ElevenLabs and gets the audio back. After that FastAPI merges the audio with the video from Higgsfield.

You do not need FastAPI for this. A simple script or a workflow in n8n is enough. We like building things reusable, which is why our own hub sits behind it.

From practice

Your own voice in three minutes.

For our comic generator we needed a voice. ElevenLabs showed us a text, which I read aloud. After about three minutes the material was enough for my own voice.

After that the voice speaks the texts that come from Notion. FastAPI merges the audio with the video from Higgsfield. The episode on it is on MacNorris TV.

  • Voice created by reading aloud a displayed text
  • Around three minutes of material were enough
  • Around 2,200 credits used including all tests
  • Audio for a 10-second video about 0.10 EUR
  • Training went nowhere at first: browser had no microphone access
  • Voice is called from FastAPI

Many try AI voices. Few use them sensibly.

Want to build something like this for yourself? We show you what is worth it and what is not. How we approach things like this is in the playbook.

Related Service

AI Consulting Without the Hype – Where AI Actually Helps

View Service

Built with ElevenLabs: AI Voices and Your Own Voice via API

AI Comic Generator with Higgsfield and ElevenLabs
Showcase

AI Comic Generator with Higgsfield and ElevenLabs

AI comic generator built from scratch: Notion as input, FastAPI as the hub, Higgsfield for video, ElevenLabs for voice. Stack, workflow, costs and what broke.

  • Higgsfield IconHiggsfield: AI Video, Soul ID and What It Costs
  • Elevenlabs IconElevenLabs: AI Voices and Your Own Voice via API
  • Notion LogoNotion: What It Does, What It Costs, When It Fits.

19 min video

Learn more