I’ll add new images whenever I feel like it.
Cindi Savagelips:
Thrall:
Queen Moonclaw:
16 Likes
Oh my gosh this is a fun idea 
Please let me know what site and prompt you’re using because this is what I got for Carnex…
Queen Beetrix…
7 Likes
What happened to poor Carnex?
Beetrix is also… Very strange…
3 Likes
This is probably as close as I can get with the generator I use.
This is the prompt for Queen Beetrix (and the generator I use):
It’s really quite decent and working as advertised. Only thing you can’t change is the image size. And inpainting would be nice.
I’ve included a few other variants that I did not like for one or another reason below.
10 Likes
Generator likes them thick thighs. 
3 Likes
3 Likes
I just tested the link I gave with three different browsers and always got my image.
Using the same prompt, order, setting and seed produces the same image with just minor differences, so somehow, you must have changed something.
Try your luck with creating a different troop and share your result 
2 Likes
Tried to generate Suna, from bright forest, to see how it handled more human figures. It’s really hard to get it to produce certain results, like the metal ring behind her head and the big red wings (it keeps putting them coming from her head, weirdly). Fun, but hard to get working.
This was as close as I got.
and that was with this prompt
Any tips on getting certain results?
3 Likes
Ah, so this runs on Stable Diffusion, too?
I don’t enjoy the concept of AI generators being able to simulate a real artist in a miniscule fraction of the time, but I will admit it’s fun to toy around with.
1 Like
Interesting toy for someone, whose drawing skill are quite… limited (I got a few old trashy MS Paint sketches, but mostly rely on what friends cobble together for me). I don’t think, this technology can or should replace real artists, but it makes for some neat basic concepts.
I wonder, if it’s limited to this realistic style, or if I can push it further into a more “cartoony” direction.
[Edit:] Still quite a bit away from how I picture myself, but we’re getting somewhere. I am partially intrigued, partially concerned, about the possibilities of modern technology…
4 Likes
I gave Perchance’s generator a rough description for Amira and got some results that are … not bad, but its full disclosure of the prompt revealed quite a few parameters which were included automatically (presumably, from the list of “styles”) rather than being explicitly input by me. Duly noted.
Getting a decent result is an iterative process: review the set of results and the prompt given, make some tweaks then try again.
Anyway, here’s one attempt at Amira (after several iterations):
The full prompt, including presets
AI Image Generator (free, no sign-up, unlimited)
anthro anthropomorphic auburn female fox assassin holding a dagger in each hand; wearing a leather corset with angular shoulder pads, leather gauntlets; a thin (green scarf) wrapped around her neck that flows like a cape; blue eyes slightly glowing, canine teeth slightly exposed, blond hair in a short ponytail; narrow vulpine ears with slightly gnawed edges and small hoop earrings digital art, masterpiece, 4k, fine details (seed:::304751242)
negative prompt:
bangs, bad photo, worst quality, bad composition, bad lighting, bad colors, small eyes, low quality, bad art, poorly drawn, deformed, bad 3D render, boring, lifeless, deformed, ugly, low resolution
5 Likes
Just change the art style. At the bottom of the list you can select “Cartoon” and see how that looks like.
As for tips: Describe the image you want to (re)create as closely as possible, check results, curse, generate again (repeat a lot of times, always making adjustments to the prompt).
If you get an image/a pose you like, add its seed to the prompt and see if you can adjust it a bit to your liking (e.g. negative prompt). Too many changes and you can get a quite different image. That’s why I would really like to have an inpainting option. It’s practically impossible to get exactly what you envisioned since you have to take the whole package. But hey, it’s free and all.
As for Suna, I kinda give up. I got a few images that I quite like, but the resemblance is … a bit distant. Too much details that would need to be changed individually. I don’t even know how that floating metal thingy
behind her is called, so I completely omitted that.
I still like all of the images, just not quite what I was aiming for.
Link to the prompt used
5 Likes
From what it seems, the 2D Disney Style option further up leads me even closer to what I search.
It might still have some issues with room awareness and item consistency, and a lot of promising attempts tag a second tail onto me, that I struggle to get rid of, but I admit, I am kind of excited (and still spooked, that we are at such a technological level by now).
4 Likes
Can someone do Frost Archer? Asking for a friend…
4 Likes
But we have Frost Archer at home:
Frost Archer at home
Prompt (no presets):
female elf archer readying her bow, blonde hair, in blueish white armor with gold trim and skirt, wearing a white hood, facing right, aiming a bow and arrow made of ice crystals, standing against a snowy mountain background, cel-shaded cartoon or anime (seed:::953639562)
Apparently, bow and arrow is a really difficult thing for Stable Diffusion to identify, even moreso than hands.
3 Likes
To be fair, it was a reference to before you even started playing GoW I think. At that time, Frost Archer was considered as one of the worst artworks for a new troop. Things have changed since then.
4 Likes
Could have told us sooner. :>
Who would have thought that a simple bow would be so difficult to produce …
I’ll stop trying now.
7 Likes
Humbly requesting Diamantina?
2 Likes