Affiliate links on Android Authority may earn us a commission. Learn more.
The best AI voice generators: Convert text to human-like speech
If you haven’t paid much attention to the latest developments in the world of generative AI, you may not realize that computers can now emulate human voices to a surprising degree. Just like how ChatGPT revolutionized the written medium, many video creators and social media personalities now rely on AI voice generators. The benefits are clear — adding a voice can make content come across as more expressive and personal. And with modern text-to-speech engines, you can fine-tune the delivery with different voices, customizable pitch, and even custom pronunciations. So without wasting any more time, here’s a list of the best AI voice generators available today.
If you’re looking for a text-to-speech product with the most diverse range of voices, you’ll be hard pressed to find one that competes with ElevenLabs. At its core, it offers AI voice generation with support for dozens of languages. But you can also go one step further with custom voices, which you can build from scratch by specifying the speaker’s gender, age, and other parameters.
ElevenLabs also allows you to clone existing voices, whether someone else’s or your own. The base tier allows you to clone a voice with audio clips as short as 60 seconds but you’ll need to upgrade to the Creator tier to create a more thorough replica of your voice. The latter costs $22 per month and also grants you roughly two hours of AI-generated generated audio. Another factor that makes ElevenLabs one of the best AI voice generators is that you can download your creations even on the free tier. You get 10,000 characters worth of audio generation per month without having to pay anything.
PlayHT claims that its AI voice generation works so well, it’s virtually impossible to distinguish from actual human speech. That certainly doesn’t hold true for all voices as a few I tested still sounded a bit robotic. But if you find the right one among the hundreds of choices, chances are that you’ll be happy with the results. PlayHT also recently showed off its new conversational text-to-voice AI model that sounds a lot more realistic, but it’s locked behind a waitlist for now.
As with most AI platforms, PlayHT requires you to subscribe to a paid plan beyond the initial free tier allowance. The minimum price of $31.2 per month certainly isn’t cheap, but the 600,000 generated words you get is a lot higher than rival platforms offer for that amount.
3. FakeYou: The best celebrity AI voice generator
If you’re looking for a celebrity AI voice generator, FakeYou performs remarkably well. The platform offers over 3,000 voices spread across categories such as television, video games, and musicians. Each voice has an associated quality rating, so you know how closely it matches the source. For example, Donald Trump’s voice had a rating of 3.5 — likely because it sounded a bit robotic. That said, the voice still matched the description and I can imagine the result would work fine for casual use cases. Arnold Schwarzenegger and Samuel L Jackson’s voice impressions are also rated higher.
FakeYou’s pricing plans are much simpler and cheaper than the competition, even though it’s one of the very few offering celebrity voices. But that’s mostly because you can only generate up to two minutes of audio at once. The cheapest paid plan, which will set you back $7 per month, grants just 30 seconds of audio and you may have to wait in a queue for each generation.
In addition to standard text-to-speech, Speechify also offers an entire AI voice studio. The latter gives you a powerful timeline-based editor for voice overs, dubbing, and transcribing. As you can see in the above screenshot of a sample project, the interface is very intuitive and easy to use.
You start off with a blank project where you can add blocks of text, each with adjustable parameters like different voices, pauses, and custom pronunciations. This means you can create an audio clip with multiple voices talking to each other in a way that sounds organic and natural. You can also add in a background audio track and corresponding imagery to preview what your final audio clip will look and sound like. Speechify also includes two official celebrity voices to choose from at the moment, namely Snoop Dogg and Gwyneth Paltrow.
Speechify Studio’s free version doesn’t let you download any audio clips, but you can get a feel for the platform and decide whether it’s worth paying for. The cheapest premium plan comes in at $288 per year, or $24 per month. Luckily, if you only want an AI generated voice to read out your emails and websites, Speechify’s text-to-speech service is quite a bit more affordable at $139 per year.
If Speechify’s AI voice studio appeals to you, you’ll also want to check out Murf.AI. It offers a similar editing interface with customizable blocks of text and sliders for pitch and narration speed. You can also add emphasis to certain words or change their pronunciation from within the editor. You get 10 minutes of audio generation as a free user, with full access to the editor and voices. Like the others on this list, you’ll have to fork over for a paid plan if you want to download the clips for your own use.
6. Tortoise-TTS: The best free AI voice generator
So far, every single AI voice generator on this list requires a monthly payment of at least a few dollars per month. Luckily, that’s not your only option if you own or have access to a powerful computer. Tortoise is regarded as the best open-source text-to-speech programs that you can download and run on your own PC with just a few commands. Be warned that converting text to natural-sounding speech is a fairly resource intensive process, so you may have to wait longer between each generation if you use slower hardware. Tortoise’s developers have put together a demo page in case you’d like to check out what it’s capable of.
Tortoise requires an Nvidia GPU or an Apple Silicon-based Mac, so it goes without saying that you’ll need a fairly recent computer. But even if you don’t meet that condition, you can use a cloud service like Google Collaboratory for free. Another open source project, titled TTS Generation WebUI, offers a one-click setup process through Google Collab that eliminates the need for any command line work whatsoever. Simply head on over to the project’s GitHub page and click on the Google Collab button to get started.
We’ve also used Google Collab in conjunction with another free project to run a chatbot in the past, in case you’re looking for a open-source alternative to ChatGPT.