What it is
Speech synthesis. You give it text, it gives you audio that mostly does not sound like a robot reading a shopping list. It also does voice cloning, dubbing into other languages, and has an API that people actually build products on.
Who buys it
Anyone producing narration at volume: course creators, publishers turning articles into audio, video teams who need a consistent voice across hundreds of clips, and developers embedding speech into their own software.
Pricing, checked 18 August 2026
Read off their own pricing page today. Verify before you commit, because credit allowances move.
- Free, 0 dollars, 10k credits a month
- Starter, 6 dollars a month, 30k credits
- Creator, 22 dollars a month, 121k credits, adds professional voice cloning
- Pro, 99 dollars a month, 600k credits
- Scale, 299 dollars a month, 1.8M credits
- Business, 990 dollars a month, 6M credits
The thing to watch is the jump from Creator to Pro. Everything is priced in credits rather than minutes, so your real cost depends on how much you regenerate, and people regenerate a lot while they are still learning the controls.
What we will test
The full nine-task protocol for this category is published on the AI voice tools page. Three of those tasks are blind listening sets, including a deliberately difficult script full of names, numbers and abbreviations, which is where most engines come apart. We also regenerate the same line three times and score the worst result, not the best.
Our commercial relationship
We earn a commission if you subscribe through our link. Their published affiliate rate is up to 22 percent, read off their own affiliate page on 18 August 2026. That rate had no bearing on this page, which contains no verdict, and it will have no bearing on the score when we publish one.