Nova Anime AM
Nova Anime Anima is Nova Anime: Anime/2.5D/3D checkpoint model for Anima
All images on example were created with diffusers with custom png tags
Rules
You cannot use the generated images for commercial use if it's not edited (or just turning it to black and white)
You can share images without any restriction if you don't monetize it
Advertising this model to outside is always welcome
Recommend Settings
Sampler: Euler a Normal
Steps: 20~30
Clip Skip: 1-2
Denoising Strength: 0.65 - 0.8
CFG Scale: 4~6
Prompt: masterpiece, best quality, score_9, score_8, score_7, year 2025, newest, highres, absurdres, very aesthetic, scenery, {Prompt}
Negative Prompts: 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
Stable fingers
Multiple subjects in image support
More art style tags support (Same level as Anima base v1.0)
Anima base v1.0 based
FAQ
Comments (19)
waiting for Kemono AM !
2.0 brings the reliable strong utility I have come to expect from Nova models. Thank you, as always, for your excellent work to bring us better and better models. A Nova model is always an easy recommendation I can be sure any user will be happy to use.
Super hyped to try the new version x)
v2.0 is VERY DAMN GOOD
waiting for moar new/updated Anima models
One question: is there a list of trained characters? Or what can we base ourselves on?
Try using Animadex
It's probably not all of them but I can guarantee that all of the presented ones can be made
Ok might be newb question but why lower denoise... what's the advantage? does it really improve or change something?
My fav Anima checkpoint so far! Would love to see your Nova 3DCG updated to Anima as well :)
Coming soon
@Crody yay, can't wait!!
@gothwomenenjoyer Uploaded!
https://civitai.red/models/2678095/nova-3dcg-am?modelVersionId=3007078
@Crody you're genuinely the goat
I am curious why the denoising strength should be less than 1.0. Any particular reason?
I do get good results starting at 0.75 (anything below that creates really weird results). But anything above 0.75 just seems to change the output entirely but not the quality.
It's for hires.fix
For i2i, something like 0.75 should work as well
For t2i, the denoise strength should be 1.0
Can somebody say please.Anima is better then illustrious?or they jsut different,because i notice that there is less lora models and loras start coming adn more push in anima
Better? not necessarily, more like a side-upgrade in my opinion. It's still lacking the Illustrious LoRA library you can find on this site.
Some things I've noticed as to why creators are switching:
1. Anima is being pushed by civitai because they have more control over its curation/ecosystem compared to SDXL and Illustrious.
2. Anima is a smaller model but just as modular and accurate if not moreso than Illustrious.
3. LoRA/Checkpoint creators are incentivized to make Anima models because they already have Illustrious datasets.
Anima has quite a few benefits - it accepts both tags and natural language prompts; you can mix and match them.
It does a marvellous job at adapting concepts and keeping characters separate - try getting illustrious to do a "round bed" and see how far you get. Same for having several distinct characters in one picture without excessive prompt bleeding unless you use reginal prompting tools.
As opposed to Qwen and Zimage, it does NSFW natively - it knows what a blowjob looks like, what shibari bondage is and what a naked man looks like, while still being able to process more complex prompts.
The current downsides are that the artstyle is very hit and miss - the base model is trained on an anime aesthetic and incosistent as fuck with the outputs, and the artist tags included range from passable to aesthetically downright awful - there's also a shitton of them, and even the browsing website helps little.
Being trained on Danbooru only, it also lacks quite a bit of furry knowledge, especially for more exotic species (I shouldn't have to prompt "horns" for a goat, but here we go...), as well as having a distinctively eastern furry aesthetic - matter of taste, sure, but it veers a bit too much into kawaii for my tastes. Fantasy races tend to look very JRPG-style indeed.
As a benefit, it's a very small model and runs on a 8GB GPU. It's also decently fast, moreso if you use the 10-step turbo LoRA.
Illustrious has existed longer, so it has more community support, LoRAs, and finetuned base models. Anima is still very young, so it doesn't have the same level of support yet, but it shows promise because it has better prompt adherence. In terms of visual quality, Illustrious is currently superior. For prompt adherence, Anima is superior, and time will tell if it surpasses Illustrious visually.
lack of lora is the main issue, it is better for prompting and better for lesbians