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
Much stable generations
Better R34 / E6 Adherences
Better lighting / colors
Anima-base v1.0 based
FAQ
Comments (7)
When I realized Nova hadn't been updated for over half a month, I just knew the next Nova 3.0 release would hit like an atomic bomb. In fact, although in my experience, having multiple characters can cause element bleeding (for example, if one character has heterochromia, the other character also ends up with a high chance of having heterochromia — though that might be down to how I wrote my prompts), I have to say this model has the best prompt adherence out of the 10+ Anima fine-tuned models I've tried. On top of that, the extra-finger problem, which I assumed was a chronic Anima flaw, has been basically solved. Not just fingers — even toes can now reliably come out as five. This is the real shot-caller of the stinky feet model world. Maybe you could put some foot pictures in the showcase to highlight this? (
English is not my native language; this passage was translated by AI, and I hope there are no misunderstandings.
The consistency has improved a lot. Version 2.5 had problems with extra fingers and limbs, but now everything is much more consistent. I liked the 3.0 update.
Hi, I love your checkpoint. Can you provide a convrot int8 variant?
Love the direction with v3, a good balance between keeping base anima diversity and consistent aesthetics.
Hello guys, when trying to use this model locally on ComfyUI, I get hit by this error from the CLIP set last layer: AttributeError: 'NoneType' object has no attribute 'clone'
I tried to remove the node but it just throws the issue down the line, been trying to fix it for more than an hour but to no success, if anyone has had the same issue and managed to fix it I would be happy to hear
I got the workflow from drag & dropping an example image into my Comfy
Based on my own experience, you can delete the "Set Clip Last Layer" node.
Since this is an Anima based checkpoint, it's not going to work in your workflow that you use for SDXL based checkpoints without some modifications. As someone else said you can try deleting the clip set last layer node, and if that doesn't work I'd suggest building a new workflow. You will need different clip and vae files as well.