ElevenLabs Review: Is It Still Great in 2024?

My ElevenLabs review can help you get a better perspective on this AI-powered platform for speech. Some consider it TTS (Text To Speech), but in my opinion, it’s so much more. Companies use it to start and complete their complex projects. It even includes the Speech to Speech option.

So, let’s have a look, shall we?

What Is ElevenLabs?

ElevenLabs is a voice AI platform that allows users to create speeches with thousands of voices in dozens of languages. It incorporates various features and one of the best AI models, making the output natural-sounding and extremely high quality.

Features

The core of my ElevenLabs review is the tour of the main features. These are the primary reasons why the company skyrocketed from a startup in 2022 to an industry leader.

Speech Synthesis

The primary feature of ElevenLabs is called Speech Synthesis. It can do two things:

• Convert text to speech
• Convert speech to speech

There are several AI models from which you can choose, regardless of the input you go for. They work best with English, but they also support 28 other languages.

Available voices are labeled with several different parameters:

• Accent
• Tone
• Potential use case.

Accents are very cool, and probably the part in which ElevenLabs voices outperform all of the competition. Besides standard American and British, there are many others you can find helpful, such as American-Italian or British-Essex.

If you’re trying to make an audio story alive, you must choose a proper Tone. It can be profound, crips, strong, or anything else you can imagine!

Finally, it’s very user-friendly, and you can choose a voice built specifically for gaming, narration, or documentaries.

And best of all:

If you aren’t satisfied with any of the premade options, you can create your own voice. This feature is available in the Free plan, so you don’t have to pay anything to test it. More about that in a bit.

Additional voice settings include:

• Stability
• Clarity + Similarity
• Style Exaggeration.

If you plan to use the same voice multiple times, you should crank up Stability almost to the maximum. This may make the voice monotonous, but the output won’t vary between generations.

Clarity + Similarity settings allow you to enhance the output but with the risk of artifacts. Keep it down if you use additional sound effects or background music.

Style Exaggeration can result in more energetic speech. On the other hand, it can also cause instability while generating, so use it with caution.

Good to know: AI Speech Classifier can identify if a voice is generated with ElevenLabs. You must upload the audio file (less than 10MB), and the system will give you the analysis. It’s limited to one minute of audio.

VoiceLab

I’ve already hinted that you can create your own voice in ElevenLabs. The VoiceLab feature allows you to clone an existing voice and quickly generate speeches with it.

Important: You can use your own voice or someone else’s with their permission.

First, it’s important to differentiate two things:

• Instant Voice Cloning (IVC)
• Professional Voice Cloning (PVC)

IVC is used for shorter samples and it doesn’t include AI model training.

The provided data must contain one speaker without background noise. The length should be over one minute.

PVC includes AI model training, and as a result, you get a hyper-realistic model. You’ll have to wait up to a month to get the result, which requires serious computing.

However, it’s well worth waiting for as you’ll get a voice identical to the one you trained it on.

Alternatively, you can choose one from the AI Voice Library. They’re publically available custom voices created by other ElevenLabs users.

The library is well-organized, and you can easily filter by:

• Category
• Gender
• Age
• Accent.

Dubbing

If you produce content in one language but want to make it available in more languages, you can use the ElevenLabs Dubbing feature.

There are many use cases for AI Dubbing, but the most popular include:

• Social Media Content
• Marketing
• E-Learning.

Unlike many other solutions out there, ElevenLabs supports multi-speaker audio and video files. This means it’s also suitable for interviews, podcasts, and other content.

To get started, you’ll have to provide input, which can be:

• MP3, MP4, WAV, and MOV file
• Youtube
• TikTok
• X (Twitter)
• Vimeo.

ElevenLabs also supports other URLs, but the listed ones will indeed work.

Once you submit the dubbing file, choose the Original and Target Languages.

There’s also an option to select the number of speakers. If you’re going to use the platform for dubbing, I recommend you select it to improve the quality of the output.

Projects

Projects are an advanced way to handle complex assignments within one environment. All of the previously mentioned features are available. The only difference is that you can use some advanced settings.

If you’re looking for a platform for creating an audio version of your book, it’s a great place to start! It can also be used when you want to include multiple voices.

You can seamlessly switch between characters, tune their voices, and regenerate them until satisfied.

There’s an option to divide the content into Chapters and bring a lot of structure to your work.

To summarize, Projects in ElevenLabs combine all the needed features to work on extensive assignments on one platform. Unlike other solutions, you don’t have to use multiple tools and waste time – they’re already set up.

Low latency API

If you’re a developer, you’re going to love this one. ElevenLabs promises an advanced text-to-speech API that responds under 400ms!

Practically, this means you can generate real-time voice in under a second. This can be suitable for chatbots, sales, and other projects.

Moreover, you can adjust the latency and optimization settings. The API documentation is well-organized, so setting it up should be easy.

There are 1000s of voices available in 29 languages!

The coolest thing is combining it with other AI models, like ChatGPT. Remember, ElevenLabs can handle turning the text into the voice, but you might need a model to generate the text first.

With the API access, you can use Instant Voice Cloning and Professional Voice Cloning as I’ve previously mentioned.

If this sounds exciting, but you aren’t a developer, check out the best Custom GPTs for coding to get started.

Overall, it was never easier to dynamically create an AI voice based on the data at hand.

How To Use ElevenLabs?

Now that you know more about the platform, let’s learn how to use it! As part of this ElevenLabs review, I’ll show you some of its main functionalities:

1. Head to ElevenLabs.io and create an account.
   If you already have an account, log in.
   
2. In the default Speech tab, select the desired input.
   It can be text or speech.
   
3. Click on the dropdown under Settings and select the model.
   I suggest Eleven Turbo v2 if you’ll use it in English.
   
4. Click on the Voice dropdown and select a suitable voice.
   You can test them by clicking the Play button.
   
5. (Optional) You can experiment with Voice Settings.
   I suggest you leave them default and play with them after generating the first output.
   
6. Add the text and hit the Generate button.
   

Here’s the result I’ve got:

Head to the Voice Library to check out how other ElevenLabs users tweaked the settings to get the desired output.

As with any AI-powered platform, it’s all about experimenting. I recommend you test as many voices as needed until you find the right one.

Pricing

The pricing is character-based but is charged as a monthly subscription (bundle). This means every time you use ElevenLabs, you also use the characters, regardless of whether you download the output.

There are six plans you can subscribe to:

• Free ($0/forever)
• Starter ($5/month)*
• Creator ($22/month)*
• Independent Publisher ($99/month)
• Growing Business ($330/month)
• Enterprise (Custom pricing)

*The first month of Starter and Creator plans offers an 80% discount. Prices are tax-exclusive, meaning you’ll likely pay more depending on where you live.

The free plan gives you 10,000 characters per month. You can also use API access and other features I’ve presented. However, you can’t use the Projects feature. Finally, you must use the ElevenLabs attribution to publish the output somewhere.

Here’s how plans compare regarding the most important parameters:

PLAN	Free	Starter	Creator	Independent Publisher	Growing Business
Characters	10,000	30,000	100,000	500,000	2,000,000
Custom Voices	3	10	30	160	660
API access	Yes	Yes	Yes	Yes	Yes
Commercial License	No	Yes	Yes	Yes	Yes
Instant Voice Cloning	No	Yes	Yes	Yes	Yes
Voice Operations Limit	Not applicable	65	95	290	1040
Professional Voice Cloning	No	No	Yes	Yes	Yes
Additional Characters (PAYG)	No	No	Yes	Yes	Yes
Analytics Dashboard	No	No	No	Yes	Yes

I’ve excluded the Enterprise plan as its limitations and pricing are custom.

Characters are renewed each month but aren’t transferred to the next one if you fail to use them.

There’s also a limitation on how many characters you can use with one request:

• 2,500 if you’re on a Free plan
• 5,000 if you’re a paying customer.

You can cancel your ElevenLabs subscription at any time. After your billing cycle ends, the system automatically downgrades your account to the Free plan.

Overall, the free plan is generous because the free characters are renewed monthly. I believe it’s more than enough to get started, and once your needs grow, you can subscribe to a paid plan.

Get ElevenLabs

ElevenLabs Review – Verdict

Is it still great in 2024? Absolutely!

Will it continue growing? Most likely, considering the number of user-generated voices.

I’d like to conclude my ElevenLabs review by saying it’s the best AI voice platform I’ve used. It holds a lot of potential. Getting the voice settings just right can be tricky, but once you do, you’ll be pretty impressed by the output.

Bonus: If you’re using Google Chrome and ChatGPT, you can try the Talk-To-ChatGPT extension. It utilizes an API connection with ElevenLabs and lets you talk to the most popular generative AI.