I have a background in AI for linguistics and translation, so I recognised most of the voices available in this solution. Coming from several sources, they range from the marginally robo-glitchy (you can tell they are generated) to the "OMG, that's a computer?!" US and UK voices tend to have an edge on other language packs, simply due to investment in vox (voice) technology going to where the market is.

What Play.ht has done is provide a workable interface over the top of cloud-based text-to-voice solutions. At this time, the solution is just that - workable. If you have content that you want to turn into voice, this solution will go a long way to making that happen for a tiny-tiny-tiny-infinitesimal fraction of the cost of pro voice-over talent.

Of course, pro voice-over talent has a quality that still can't be beat. (Though budget voice-over talent is probably worse than the AI voices from this pack!) So if you are in a blogging niche can't afford pro voice-over talent, then this solution may just what you need.

I do understand the costs associated with this type of AI processing. It ain't cheap to run the compute needed for this stuff. So the limited credits per month is still a good deal.

I found the chat-based customer support to be on-the-ball. Very impressed.

Now the downside. I mentioned above that the application is 'workable.' It is. It does the job of turning blogs into voice. However, there is a lot of room for improvement in the UX. For example, finding the dashboard to fine-tune your text to speech is a little tricky and there is no 'save work in progress' feature. If you accidentally with the ESC key while editing, FATANG!, the work area closes, your work is lost and you need to start from scratch.

So the product feels a tad underbaked. But that is what AppSumo is all about! Getting access to the next great thing at prices that can't be beat.

Despite the UX issues, I am giving this product five tacos. Four for a solid app that does exactly what it promises, plus one more for excellent customer support.

You will need to be committed to getting value from this package. Good text-to-voice does not just happen. You need to review and tweek the work. For example, the company name IBRS needs to be rec-oded to eye-bee-are-es for accurate pronuction with some voices. Technical terms will need similar treatment. You can expect to spend 10-30 minutes per blog getting the text-to-voice just right.