IS THIS YOUR FURSONA?

We’ve been playing with AI art generation. Like everyone else in the latter half of 2022, really.

Naturally, we’ve been trying to generate more art of ourselves, using Stable Diffusion, on our own computer, cranking out things and tweaking prompts and trying to learn how to tease out of it at least something that’s editably close to what we look like.

What have we learned? So far, we’ve learned it really likes Skye much better than the rest of us, and most AIs (we have tried Dall-e 2, Midjourney and Stable Diffusion) have absolutely no idea what a nekomimi, catboy or heck, ‘boy with cat ears’ (even if it’s ‘cat ears on his head’!) is. It’s super hard to get an image out of any of them for Rhodri! Not that it’s easy to get one for Finn, either…

So, what happens when you’re testing prompts and you have no idea what will work, and you crank out tons of images trying to get anything close to what you were looking for?

You get a lot of failures. (A LOT!)

However, you also end up with some output that isn’t you, it isn’t what you need, or want, but it’s weirdly… really good, or at least it looks like it wouldn’t take a lot of editing to fix up a bit.

We looked at some of these good, but not good for us images and thought….

“Well, that’s a shame. This is someone’s fursona.”

(Please note that some of them are not furry, because they are human or mostly human, and were usually – not always – attempts to get nekomimis. So it’s also ‘is this your avatar’ in general, but.. it’s mostly furry.)

Then, when we found a good prompt for getting some kind of furry result.. well, we started cranking out species that have nothing to do with us. Sometimes, just to see if it understood the species. Sometimes, to experiment with prompt crafting, to zero in on things. This just generated more images that.. well, they’re probably somebody’s fursona.

So we decided to let people have them. Free! Here are the terms:

YES, THE IMAGES IN THE GALLERY ARE COMPLETELY FREE!

But, if you would like to thank us for taking the time to post (and categorize!) these for you, and for using our web hosting and bandwidth, you can either join our Patreon, or give us a donation on Ko-fi! You don’t need to, but we really appreciate anyone who does contribute!

NO, WE ARE NOT OFFERING IMAGE EDITING, OR CUSTOM IMAGE GENERATION SERVICES, THE IMAGES IN THIS GALLERY ARE PRESENTED AS-IS AND MOST WILL NEED EDITING!

These are images that were created as a ‘side effect’ of our own image generating projects, or ones we did just for kicks, to see if we could. They are presented and offered as-is, with no additional services attached. Many, if not all, of the images in the gallery, have some kind of ‘defect’. Sometimes this is something that can be fixed with mere cropping (most of the hands), sometimes it will require some additional editing to get them to look even ‘correct’, much less looking like your fursona. All images were selected with the expectation that the user would have some basic image editing knowledge, or be able to find someone else who can edit for them. We may not have your species and gender, or not have them in the style you like. We do not take requests, unless we are streaming the process specifically for that purpose. You may see a lot of the same kind of thing, usually as side effects of attempting to zero in on images of us. However, there is variety, because we run what’s fun (and ‘fun’ sometimes means ‘hey, let’s see if it likes this species’), and collect what looked good, and put it up here. If you would like to create your own AI images with Stable Diffusion, it is an open source project with many ways to install and use it, we suggest you look it up… if Rhodri can figure out a way to install and run it, you can figure out how to install it and run it, too. (Okay, I had some help, but seriously, if I can do it, you can do it too. – Rhodri)

YOU DO NOT HAVE EXCLUSIVE RIGHTS TO ANY IMAGE FROM THE GALLERY

By using one of these generated images, you are accepting that you do not have exclusive rights to this image. We highly suggest that whatever image you take from the gallery, you edit to some degree – color, cropping, something, anything – to make it more unique to you, because the base image is available for anybody to use. That means you and someone else could be using the exact same image for both of your avatar images, even on the same website, and both they and you are within your rights to be using it, whether either of you altered it or not. Think of it a little bit like a YCH, only the part where you turn the base into your character is the part that you do, we’re just giving you a very detailed base to start with.

YOU CANNOT CLAIM COPYRIGHT OF THE GALLERY IMAGES AND YOU MUST DISCLOSE THE USE OF STABLE DIFFUSION AI IN CREATION OF THE GALLERY IMAGE THAT YOU USE

This is related to the above. You cannot claim that someone else is using ‘your’ copyrighted image. They belong to everyone. We are releasing the gallery images under Creative Commons Attribution-Sharealike, you may produce derivative works (that is, you can edit the images as you like – and you should!), you can even use them for commercial purposes… however, everyone else can do that, too, and any works you produce from it must also be licensed the same way. You cannot trademark the images, and we would definitely not suggest that you use them, or your derivative work, as a logo, or branding, because of this!

For attribution, you do not need to credit Pryderi System (you can mention us if you like, but it’s not necessary or expected), but you absolutely are required to credit Stable Diffusion AI, and you cannot claim to have created the work in its entirety (even if you edit the heck out of it and do a lot of painting over) all on your own. You must disclose that the base image was created by the AI.

YOU MUST NOT EDIT FOR OR USE THE GALLERY IMAGES IN HATE SPEECH

Per the user terms for Stable Diffusion AI, you cannot use the images to promote hate such as anti-semitism, racism, misogyny, homophobia, transphobia, etc. You cannot edit hateful imagery onto the base image (no adding Nazi imagery, for example), and you cannot use them in a product that is promoting hate or encouraging violence toward others.

OTHERWISE, GO NUTS AND ENJOY THEM!

Seriously! Have fun, make avatars for sites, print out con badges, use them as a base for larger paintings, make a tarot deck, wallpaper your room, just do something with them, because we have them and don’t need them, but somebody does! Are you that person? Do you need an avatar? Do you see your fursona here? Do you see a new fursona here? Let it inspire you, write a story about the character you found, draw more art, let it be a beginning of something, or just a simple visual aid. Just follow the very simple rules (attribute to Stable Diffusion, no hate, no telling others they can’t use the same image, basically), and use your imagination!

DISCLAIMERS:

Pryderi System does not warrant or guarantee in any way that the images contained in this gallery are suitable for your particular use, nor do they make any guarantee or promise of the legality or continued legality of any use you put them to. The world of AI art is very new, and the legal landscape around it is likely to change quickly in the next few years, and no promise is implied or given that there will not be any legal restrictions on these images at any time, and by using one of the gallery images, you are agreeing that it is your responsibility to keep up with the law as it relates to your use of the image, Pryderi System has no responsibility to inform you in any way of changes to the general terms of use of AI generated images.

We do guarantee that every single one of these gallery images are direct, unedited output from Stable Diffusion AI, no matter how that ends up affecting them legally (or not) in the future. However, there is a possibility that some of them may be close enough to a specific piece of a human artist’s work to cause questions, and we have not reverse image-searched these to look for potential issues. The images are presented as-is – before you use them, if you wish to avoid any potential problems, you are expected to do any due-diligence you feel is required for your particular use-case for the image. By using an image, you are agreeing that it is your responsibility, not the responsibility of Pryderi System.

All disclaimers, offers and terms of service apply only to the images in the Fursona Gallery, and not to any images used in the layout, display and decoration of the website, which are from a mix of sources (including Stable Diffusion, but not exclusively), and are not offered for public use.

MISSION STATEMENT:

We understand that people are worried about AI art taking over for real, live artists and depriving them of work. We also understand there’s worry in the furry community about this, as many people make, or supplement, their income, by producing art. We do not believe that it’s likely to take over entirely any time soon, after our increasingly extensive hobby-level exploration of it, there are a lot of things that AI is very bad at. You should definitely not use the outputs on this website as evidence of its frequent success, because this is a curated selection of the most successful outputs, this does not represent the many, many failures that went along with them! We do understand that this site might be worrying to some artists, we’re artists ourselves, though not in the drawing/painting realm, so very much. We can understand the worry, though, absolutely, so given that, we felt a mission statement, or statement of our intention with this site, was worth doing.

In no way do we intend for these images to be complete replacements for commissioned art. Our vision for the uses of this art is as a supplement to commissioned art. We have no interest in making any attempt to control what users actually do with the art (within the bounds of the terms, which are very open), but how we envision these ‘fursona images’ being used are in situations like:

You have no art of your fursona yet, but you’re finding it hard to get commissions without at least one visual reference. (We’ve been there!)
You just came up with a character, you’ve got a whole history and personality and general species, but dang, you just can’t exactly zero in on the appearance.
You just want a quick icon or avatar, and let’s be honest, you don’t have the cash to pay someone fairly to do the art for you right now. (We get it. Just consider using it as a reference for a commission later, if you stick with the character.)
You’re writing a story and you want a visual reference for you or your readers for what the characters look like (we see this as supplementing the art of the writing).
You’re trying to come up with a character, and you’re just kind of stuck. You want them to be furry, but.. everything just feels boring and you need an idea.

In general, we envision these images as inspiration, as jumpstarts to more creative work, whether that’s for artists to make new images of the same character, writers to write stories, or people to RP as characters. We don’t feel that our intent is in any way to replace artists, but to create new opportunities for people, in general, to be creative. The landscape of things may change, but we don’t intend for our part in it, here, to destroy it, only be part of the changing ecosystem, while adding creative fuel – we hope people will use that fuel to go on to create more works of art, whatever form that art may take!

ABOUT THE FILTER CATEGORIES/HOW TO

You may disagree with tags. You may find them too general. This is what you get. This is how far we’re willing to tag, and they’re tagged according to what we see in them. You may have to dig around a bit, and you may want to edit something that we say is one thing but you see as another, and that’s totally okay!

If you’re looking for something that might pass as an unusual hybrid, something real rare and unusual, or something new and interesting, try the “Unsure” category. That’s where we put anything that isn’t what we asked it for (at least, not recognizably), but looks cute, interesting or at least ‘well rendered’ in that it looks like it could be something, we just don’t know what that ‘something’ is, species-wise. So for the more eclectic sorts, that might be a gold mine! … Or not, depending on how particular and/or lucky you are. If we can tell what it is (no guarantee that you can tell what it is if you don’t know what we prompted it for), but it’s not common, it’s more likely in the ‘misc’ categories for small (dog or smaller) or large animal categories when it just didn’t fit well in one of the others.

ALL IMAGES ARE PRESENTED AT THE FULL SIZE IN WHICH THEY WERE RENDERED. BECAUSE OF THIS, WE HIGHLY SUGGEST YOU VIEW THIS ON A DESKTOP OR TABLET, AND NOT A PHONE. SIMPLY RIGHT CLICK AND ‘SAVE AS’ TO DOWNLOAD THEM.

If you wish to have larger versions, there are numerous ways to upscale images, including using Stable Diffusion itself – just look it up! What you see here is what you get, when testing prompts, we generally run them no larger than 728px in any one dimension, though you may, very occasionally, find one that’s larger than that.