Steve.ai is still a work-in-progress but it is fairly important to know what you are buying. I've gone ahead and optimistically given this 5-tacos because it looks like Ruban and team are working hard to address the CONS (below). In other words: I'm rating it on what it is going to likely be.

Some of the reviews have expressed frustration by "surprise" charges from premium or elite media. You need to be careful when you are "swapping" video/images to select "free" when you are doing searches. Hopefully Ruban will improve this process.

That being said, Steve.ai shows tremendous promise. I have Lumen5 and the biggest advantage of Steve.ai is the text-to-speech option. This, and this alone, beats the pants off of Lumen5. Steve.ai needs some work so that you can change the text in the slides while not impacting the speech.

PROS:

Steve.AI is extremely simple to use. You can paste a blog post's URL and let Steve's AI engine do the rest. You can also write your own scripts (subject to the caveats that follow) and it will try to fit a video around what you write.

The AI voice narration is decent (but could be improved). You simply type in your script and it turns your script into video scenes. Like any other text-to-speech product, though, you may find it mispronounces more esoteric things and there is no real way to correct it that we've found (more on this below).

You can upload your own videos/images to include in your scenes. This is super handy if you're making a product video, touring a facility or showing a "how-to" video. Also, if you can't find "free" stock, this offers an alternative.

You can make minor edits to the semi-finished product by hand in the "workspace" to fine-tune what goes in the video. So if you want text to look a different way, a different slide layout, or a different piece of media it is pretty easy and intuitive to make that happen.

CONS:

Steve.ai seems to be somewhat platform dependent on the Chrome browser; performance can be heavily degraded in Firefox (others?). [Side note: disabling Duck Duck Go privacy essentials and AdBlock seemed to help on Firefox]

If you have something difficult for the voiceover engine to pronounce, there doesn't appear to be a way to work around it. For instance, I teach a course "ITOM 2308". Steve.ai renders this "it-um two-thousand three-hundred eight." I can change the script to "I-Tom 23 08" and it speaks it correctly but the slide text then looks weird--and there's no way to fix this.

If you select "Music" for your video, you can't control which slides get it and which slides don't. You can't turn the music off on just one slide. So you can't have a silent "transition" slide from one topic to another. You either get music everywhere or nowhere.

It is easy to accidentally insert $$$$$s of premium or elite stock clips into your video. When you click "publish," and get the ugly surprise that you owe money, it can be confusing to see exactly which slides cost extra and which ones do