To use ElevenLabs, sign up at elevenlabs.io (you start on the free plan), select Text to Speech in the sidebar, choose a voice, paste or type your script and click Generate speech. Download the file as an MP3 or WAV. On the free plan you can't use the audio commercially and must credit ElevenLabs if you publish it.
ElevenLabs is an AI voice platform. Most people use it for text to speech, turning a script into natural-sounding narration for videos, podcasts, ads or training. It also clones voices, dubs videos into other languages, makes sound effects and music, and cleans background noise out of recordings. Here's how to get started, and the rules on the free plan that catch people out.
Step 1: Create an account
Sign up at elevenlabs.io. ElevenLabs says you're automatically put on the free plan when you sign up. It sometimes runs first-month discounts on paid plans, which you'll see on your Subscription page if there's one running.
Step 2: Make your first voiceover
- Log in and select Text to Speech in the sidebar.
- Choose a voice. You can browse the Voice Library or use one of the default voices.
- Paste or type your script into the text box.
- Click Generate speech and listen to the result.
- Click the download button at the bottom right to save it.
If you need a voice for social video and don't know where to begin, ElevenLabs names some of its most popular voices online: Brian, Liam and Adam among the male voices, and Jessica, Matilda and Sarah among the female ones.
Step 3: Download in the right format
Every file you make is saved in your history. In Text to Speech, open the history tab in the panel on the right (or the history icon above the Generate speech button on a narrow screen) and click the download icon. You can choose MP3 at 128kbps or WAV, or click Advanced for MP3 at 192kbps or 256kbps, M4A or FLAC.
For video and podcasts, MP3 at 192kbps is a good everyday choice. Use WAV if you're going to edit the audio further.
Step 4: Clone your own voice
ElevenLabs offers two kinds of voice clone, and they suit different jobs.
| Type | Audio you need | What to expect |
|---|---|---|
| Instant Voice Cloning | 1 to 2 minutes of good audio | Quick to make, ready almost straight away |
| Professional Voice Cloning | 30 to 180 minutes of good audio | A trained model of your voice; fine-tuning usually takes 3 to 6 hours |
The detail most people get wrong is quantity. For Instant Voice Cloning, ElevenLabs says more than two or three minutes of audio gives little improvement and can even make the clone less stable. Quality matters far more: use a clean recording with no background noise, room echo or other speakers, at a steady volume and without long silences. It recommends MP3 at 192kbps or above, and says uncompressed WAV files usually don't improve the result.
Professional clones need you to verify your voice by reading lines into your microphone. If you fail every attempt, you can wait 24 hours and try again.
How credits work
Everything in ElevenLabs draws from one shared monthly credit pool. Its pricing page says, as of October 2026, that text to speech costs about one credit per character on most models, so a 1,000-character script uses around 1,000 credits. Other tools cost more: sound effects about 200 credits per generation, voice changing and voice isolation about 1,000 credits per minute, and dubbing about 2,000 credits per minute.
| Plan (US dollars, monthly) | Credits a month |
|---|---|
| Free, $0 | 10,000 |
| Starter, $6 | 30,000 |
| Creator, $22 | 121,000 |
| Pro, $99 | 600,000 |
On paid plans, unused credits roll over for up to two months while you keep an active subscription, but ElevenLabs says downgrading or cancelling forfeits them.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
Other ElevenLabs tools worth trying
- Voice Isolator strips background noise out of a recording. Upload a file or record one, then click Isolate voice.
- Studio is ElevenLabs' tool for long-form audio such as audiobooks, and it's available on every plan, including free.
- Create a podcast (GenFM), inside Studio, needs a paid subscription.
Can you use ElevenLabs audio commercially?
This is the most important rule to know before you publish anything. ElevenLabs says the free plan doesn't include a commercial licence and can't be used for any commercial purpose. If you publish content made on the free plan, or without being signed in, you must attribute it to ElevenLabs by putting elevenlabs.io or 11.ai in the title. All paid plans include a commercial licence, except for Beta Services.
So if the voiceover is for a client video, an ad or anything that promotes your business, you need a paid plan. Content made in the separate ElevenReader app is for personal listening only and can't be exported or used commercially.
If the voiceover is for an ad, our AI video ad generators guide covers tools that also create the footage and AI actors.
My advice
Use the free plan to test voices and scripts, then move to Starter the day you need audio for your business, because that's when the commercial licence starts. Record your clone sample once, in a quiet room, rather than uploading lots of clips. If you're still choosing a tool, my guide on whether ElevenLabs is free goes deeper on the free plan, and if you want to animate a photo with your new voice, see where to make a photo talk with AI for free.
Frequently asked questions
Is ElevenLabs free?
Yes, there's a free plan, and you're placed on it when you sign up. As of October 2026, ElevenLabs lists 10,000 credits a month on the free plan, but it doesn't include a commercial licence.
Can I use ElevenLabs free plan audio on YouTube?
Only if it isn't commercial, and you must attribute it by including elevenlabs.io or 11.ai in the title. Commercial use needs a paid plan.
How much audio do I need to clone my voice in ElevenLabs?
ElevenLabs recommends 1 to 2 minutes of good audio for Instant Voice Cloning and 30 to 180 minutes for Professional Voice Cloning. Clean audio matters more than length.
How long does a Professional Voice Clone take?
ElevenLabs says fine-tuning usually takes 3 to 6 hours, depending on the queue, and you get an email when it's ready.
What file formats can I download from ElevenLabs?
MP3 at 128kbps or WAV, plus MP3 at 192kbps or 256kbps, M4A or FLAC under Advanced.
