While the concept is nothing new I’ve been a supporter of BrainFever since the beginning so I’d much rather go with this app rather than a competitor. So far it works pretty well. I had a little confusion early on with the perimeters but figured it out. This is just a suggestion/wish list since it is very early in the apps progression but if there was a choice of angle or perspective then it might be easier to mix a few pieces into a larger project. For example if I want a skull and I want it looking sideways or head on I can’t change that so it might be hard to use it with my other images. The other thing is the resolution. The max I’ve seen is 1024 X 1024. I’m sure this will increase over time. Again these are just ideas. Great start so far. Good luck to the developer! 👊🏼

I was able to create some great fantasy images with Dalle3. Flux model does not produce the type of images that I had been creating most of the prompt details do not even get used and its disappointing but it is the best of the models for adding text. The Flux model creates images that I could simply create by creating layered images in a photo editor which would produce better results and landscape orientation is ignored and a square image is produced. If you want realistic images then any of the models will work but most images created will have minor flaws and most of the flaws are easy to spot this is why I prefer to create Fantasy type images with AI as most flaws wont matter do to the image concept. The images are usable but not impressive.

I have been searching and searching for an ai generator app that allows for you to expand/out paint of the image. But I either can’t find one(for free) or they don’t give me the results I wanted. So glad to finally have one that actually follows the correct textures of the image. I would recommend that there be a way to gain more tokens of usage without having to pay though. But for what it is, the image expander is the best thing here so far.

I'm a little calmer today than I was when I wrote that first review, since when I've tried out just about every AI text-to-image generator on the Web and learned a few important things. First of all, if you use the DALL 3 model instead of the crappy Leonardo one that acts as the app's default right out of the box, you'll get a much more accurate image every time. Second, as AI apps go, this one is far from the only piece of junk out there, just one of the worst when it comes to handling prompts, and agonizingly slow. I'm writing a book that contains a goodly amount of material on the Stone Age, including the time before people killed animals for meat or wore ahnimal skins. I wanted a depiction of a man from that era spearing an eland; the app turned my specific instruction 'spear' into a bow and arrow - not once, but every time (about a dozen wasted attempts) - but then, so did most other apps. I wanted a depiction of people standing around their first fire, long before they had anything much to cover their bodies with; again, according to the app, therev was never such a time, or at least it didn;t know about it. I wanted a depiction of men and women from that era standing on one side of a campfire listening to a would-be leader speaking; the app put everyone on chairs and gave one of the women three legs. I asked for a man and woman talking, and every time got two obvious, heavily bearded men instead. And all that for $20 a month; when I found that most such apps - no, most BETTER apps - ran about $10, and some cost as little as $6! So instead, I cancelled this piece of junk and right now am trying to choose between Wowzer, Arcana, Open Art, and AI Tubo, all of which had no trouble letting me play with their app a little online to check out its performance, and no trouble depicting naked cave men and women standing around a campfire, people hunting with spears, men and women talking, etc., and then, ranged from $6 month-by-month to $10, or in one case, $15. Hope all this helps.