Nova Furry AM
A 2d/2.5d furry checkpoint model that can have great details on any type of furs, scales and feathers, which aims to be Anima version of Nova Furry XL
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
Steps: 20-30
CFG Scale: 4-6
Prompt: masterpiece, best quality, score_9, score_8, score_7, year 2025, newest, highres, absurdres, very aesthetic, scenery, furry, anthro, {Prompt}
Negative Prompts: human, smooth skin, worst quality, low quality, early, old, score_1, score_2, score_3, cartoon, graphic, painting, crayon, graphite, abstract, glitch, deformed, mutated, ugly, disfigured, long body, bad anatomy, bad hands, missing fingers, extra fingers, extra digits, fewer digits, cropped, very displeasing, artist name, blurry, jpeg artifacts, lowres, censor
Description
Changes
Make the style closer to the Furry XL, much western feel
Better adherence for E621 tags
Better colors
Change the base into Anima base v1.0
Anima base v1.0 based
FAQ
Comments (19)
Im gonna use this checkpont with Anima Furry Domain Lora to create a super futa furry 馃ぃ
was waiting for that update x) ! hope that one day the mobian checkpoint and pokemon one going to be ported to anima too ! i think those would be nice on anima
Any plans to publish or make a list of furry artists for the AM series? Ive noticed anima still has some trouble generating anthro characters without lora. what are the changes with this from AM1?
The snouts are too long and thin. Not cute.
Anyway, thanks for model updates. In my opinion, these are the best furry models.
Opinions are fine. But wrong ones should be kept to one's self. :)
Should have omitted the first line and just posted the second.
@KnottySergal,聽then you should have followed your own advice.
I actually like the thin snouts. It's a different look than the other furry checkpoints/LoRAs out there, and the variety is nice.
That said, if it were controllable so you could specify it or not it would be even better.
May I ask about the number and resolution of the training set used for training? I have observed a significant improvement in the details and text stability of the nova model
Anima cannot control the number of fingers and toes. Whether on the Anima base or on this model, it cannot reliably generate characters with 4 or 3 fingers. Am I the only one with this problem?
That's the thing I also noticed as well
Anima is still developing in that kind of area (especially with furry knowledges) so in the future versions, I hope to see those
Some characters however could generate with fewer digits (ex. Lucario / Judy gen) but as far as I tested, 3 toes is not the feature for the newest model yet
Prompt adherence in V2.0 is a tiny bit lower. I'm seeing some prompt bleeding across characters in prompts that were 100% stable in V1.0. Might be the natural language captions from Anima being overridden and buried?
Certain concepts are missing but I've seen the same in the og Nova Furry so it's just a blank spot in the e621 dataset.
This model lacks (has no or poor knowledge of) a lot of important (for a furry oriented model) stuff you easily find on e621, for example : some characters like Cynder (Spyro), anatomy like cloaca or teats, general concepts like plantigrade/biped,.. I suppose it is because it is an early version and it will improve in the next ones. Great model anyway, as usual, thanks! Looking forward to new versions.
Same for sheath, and probably other degenerate things I haven't checked yet, I'll update this comment if needed. Also maybe other people can generate all this very well and it's just me and my workflow or I'm bad at prompting, in this case I apology.
I've noticed when doing certain sex scenes (especially doggystyle), the model seems to really lean towards anal, even tho i dont prompt it and even if i put it in negative.
@Crody can you please also upload this model to PixAI? 馃檹
I'd like to if only they would allow Anima models
And unfortunately, since they already have their own DiT base model, I think the probability of doing this is near to none
@Crody聽 i have uploaded anima models before on there it works fine
@123inactace聽It's not available as far as I searched
@Crody聽it seems ilke PixAI added an update that lets people upload "DiT-architecture models and LoRAs." Can you check if this model can be added now?