Anima version is available!
Nova Furry XL
A 2d/2.5d furry checkpoint model that can have great details on any type of furs, scales and feathers, which aims to be XL-fied Peaki Furry
Rules
You cannot use the generated images for commercial use if it's not edited (or just turning it to black and white)
Advertising is always welcome
Recommended settings:
Sampler: Euler a for best performance
Steps: 20-50 steps
(Pony)
CFG Scale: 6-7
Prompt: score_9, score_8_up, score_7_up, score_6_up, score_5_up, score_4_up, source_furry, BREAK, furry, anthro, detailed face eyes and fur, {Animal Type}
Negative Prompt: blurry, jpeg artifacts, username, watermark, signature, normal quality, worst quality, large head, low quality, text, error, missing fingers, extra digits, fewer digits, bad eye
(Illustrious)
CFG Scale: 3-5
Prompt: masterpiece, best quality, amazing quality, very aesthetic, high resolution, ultra-detailed, absurdres, newest, scenery, furry, anthro, {Prompt}, BREAK, depth of field, detailed fluffy fur, volumetric lighting
Negative Prompt: human, multiple tails, modern, recent, old, oldest, graphic, cartoon, text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, lowres, bad anatomy, bad hands, missing fingers, extra digits, fewer digits, cropped, very displeasing, (worst quality, bad quality:1.2), bad anatomy, sketch, jpeg artifacts, signature, watermark, username, simple background, conjoined, bad ai-generated
Description
Enhanced the details and lightings to get the both A type and B type's elements
To get the shading of realistic shading, try using realistic, photorealistic
NoobAI EPS v1.1 + Illustrious v2.0-stable DARE applied
FAQ
Comments (32)
@Crody I know the PixAI transfer service from Civitai is down right now, but can you please upload it manually? I really like your models but can really only use them on PixAI because of the constraints here. Thank you.
@Crody THANK YOU!
V13 is pretty much fantastic. reminds me of when I use to 50/50 blend v8B + v8A. It reminds me so much of that effect, but this is clearly a more polished version.
Oooh. I wanna make a suggestion. After playing around with your models for so long with merging.
V14 would look really good if you built Base off V10 + one of the older ver B models.
I don't like V13. The colors became faded, and there were more artifacts in the eyes. V12, on the other hand, had more contrast and the eyes were more detailed.
Dude, your model is amazing! I've been using it for several versions now, and it's the best for creating furries.
Mine were broken... few attempts at first generate beautiful images, suddenly it became fried & destroyed --- Does it have an error in my prompt? Or the generator being whacky right now?
masterpiece, best quality, amazing quality, very aesthetic, 4k, high resolution, ultra-detailed, absurdres, newest, scenery, 1girl, furry, anthro, solo, white wolf girl, white fur, long hair, multicolored hair, (gradient hair, white hair and black hair:1.2), black tip ears, floating hair, ponytail hair, (purple leaf hairtie:1.3), black visor hat, purple eyes, very large eyes, one raised eyebrow, (sparkles:1.4), neck tuft, facing to viewer, looking at viewer, cheerful, long snout, mouth open, smile, sharp teeth, silver choker with a purple gem, medium breasts, white collared shirt, long sleeved, coffee logo, holding coffee cup with lid high, black ribbon, black slim watch, standing, arching, indoor, cafe, hanging menu boards, upper body, at eye level, close-up, dutch angle, dappled light, BREAK, cinematic, detailed eyes, detailed fluffy fur, depth of field, volumetric lighting
NEG
3d, rendered, human, multiple tails, old, oldest, cartoon, graphic, text, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, lowres, bad anatomy, bad hands, missing fingers, extra digits, fewer digits, very displeasing, (worst quality, bad quality:1.2), sketch, jpeg artifacts, signature, watermark, username, simple background, conjoined, bad ai-generated
Please have a try
@markminxd229 Tried on PixAI and it didn't have any signs of over fry
If you’re using the onsite-generator, I think it's experiencing some glitches that comes from high usages
I'd recommend using either PixAI or SeaArtAI instead which you can see on the description
Is there a list of all the characters it can produce?
There's no official documentation about character tags but I can say that characters that has the tag and made into e621 or danbooru before Spring 2025 works really well
Some characters from Summer would also work but I cannot guarantee
V12 is better than V13
I originally tried this checkpoint for its reputation in the furry domain—but I quickly realized it’s much more versatile than expected. Fantasy and sci-fi themes are handled exceptionally well, with effects that feel vivid, immersive, and technically refined.
What really impressed me is how well it captures atmosphere and detail. Whether you're aiming for magical landscapes, futuristic characters, or surreal compositions, this checkpoint delivers consistently strong results.
I'm genuinely glad I gave it a shot. It’s now firmly one of my favorites and a regular part of my workflow.
V13 seems like a massive downgrade from V12.
If you're doing solo characters, v13 is still quite nice, but overall prompt recognition, color, anatomy, and genning multiple characters is all noticeably worse. Pretty much all the models are good at solo gens, so hard to really give any points here over V12.
Regardless, still the best model and it's not even close, but would highly recommend sticking to V12.
I've decided creating IL v13.5 based on the feedbacks I got
Current-v3: https://imgur.com/a/eWEYAym
Current-v2: https://imgur.com/a/Fag0Zik
Current-v1: https://imgur.com/a/g0Qbp83
If there's suggestion, please reply to this comment (except expanding the dataset)
Below is the memo for myself
• Coloring must be the middle of v12 and v13 and needs more contrast
• Multiple subject improvements
• Improve the eye
Here's the current v13.5 look (beta stage)
https://imgur.com/a/g0Qbp83
The second version features much better anatomy
https://imgur.com/a/Fag0Zik
Third version now is adjusted to resolve the issues revolving details such as fingers and patterns
https://imgur.com/a/eWEYAym
Why not ? But please don't remove V12
and v13.5 needs more contrast and skin details than V13
@SupergirlX I won't delete any of released versions
Checkout third version of the v13.5, it has much more contrast than v13 and fine details for fur / skin
Unfortunately I feel like this series is getting progressively more "sloppified" with each iteration. The base style is way too strong and it makes style loras really hard to use with it.
Backgrounds are also very very noisy.
And if I may be blunt, how the heck does a furry model not know what a sheath is?
I want to make the model that doesn't require great amount of LoRAs to create great images
Background noise can be prevented by removing scenery from positive and (particles, adversarial noise:1.2) to the negative prompt
Sheath is the biggest problem I have for this model for a long time
Best way to do is to use these series of LoRAs
https://civitai.com/models/1173463/furry-anatomy-options-il-nsfw
@Crody I understand wanting it to be as lora light as possible, but the model in its current state doesn't allow for the mixing of style loras as much as it used to, which is my primary use-case for character art.
@Surveillance For the next version (v13.5), I'm trying to fix those issues as well
I don't know wether the look suits to your liking but here's the current images
@Crody They look very nice and I do look forward to how the model evolves both visually and functionally in the future. :]
This sheath problem exists since V1.0 and it also annoys me. But I could solve it with a Lora. Leaving that aside, I am happy with V12! If the look/colors are a bit aggressive to you, just experimenting with merging with NovaAnimalV8. The resulting hybrid look is great.
NovaFurry for me was always about using slop default 2.5D shiny style that looks good, but not original because everything was flooded with that style. When I don’t want to use default 2.5D shiny style, I just switch to another merge. For example, you can try StableMondAI-SDG or Wahtastic Mix. They also both VPRED (better colors) and doesn’t require insane artist weight to make it work like it happens on NovaFurry.
Based on the v13.5-beta3's reactions, I created final release version using it as base
This version will be available around Nov 11th 0:00 AM GMT
V13 is okay compared to V12. imo it’s less flexible in comparison, and you need to tone down the CFG scale or the sampling steps due to it being more detailed. if you want consistency, realism, or detailed/dramatic lighting emphasizing background scenery then V13 is for you. if you prefer flexibility, varied results, and character expression, then V12 is for you
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.