(22/10/2026): v3.0 should be the final version for Anima. Next: MOSS Image (aka. The greatest anime model that has ever been released.). Here.
RDBT [Anima]
A general finetuned model, from a personal all-in-one dataset, mainly for digital art and drawings, focusing on lighting and textures, from wallpaper level digital art to 5-min sketch.
Dataset contains ~10k images, handpicked by professional amateur artists. The criteria for evaluation are not visual appeal, but the drawing technique and the level of completion. Captioned by LLM.
Some cover images look ridiculous; they're just for demonstration purposes for corner cases.
Update log and version info: link.
Useful links:
RDBT | Digital Art: My merged 16-step ckpt optimized for digital anime art. 4x faster. Easy to prompt.
RDBT | Original LoRA: LoRA files for advanced users.
16-step Acc LoRA: Style friendly, less AI slop, compared to the official Turbo.
RDBT | 160KE6 Furry: Based on 160kE6Anima by @garaggge. 160kE6Anima "was trained on around 160k images from e621". It has extended knowledge and 3k+ style tags. Check its model page above for more info (list of styles, etc.).
Settings:
Base version: CFG 4 and steps 24+. Euler.
Distilled version(8s, 16s, discontinued): CFG: 1. Steps: 8 or 16. Euler.
Prompt:
Always specify @style in prompt, or use a style LoRA. This model does NOT provide overfitted default style (ignore prompt and is always active). You need to specify style in prompt, or use a style LoRA. This is a feature, not a bug. The dataset is big enough and model is not overfitted. Otherwise, you will get random/mixed style.
Quality tags:
You don't need quality tags. The dataset has higher quality than "masterpiece". Thus quality tags don't have effects. Omitting those redundant tokens allows LLM to pay more attention on other words.
Sharing merges using this model is not allowed.
This mode is a free and it will always be free. This "restriction" won't affect anyone. It's only aimed at those who steal others' models to sell.
Known model thieves: NukeA.I (selling this model behind paywall on tensorart).
Because he is such a jerk I wrote a story about it. Also contains a guide for trainers about "how to bake special trigger word into your model".
Description
FAQ
Comments (13)
this poster claims to have made the nag work with anima using turbo and 1 cfg but with your model at least , it changes the output too much. Can you take a look at it , maybe we need specific settings for it to work with your model ? https://www.reddit.com/r/StableDiffusion/comments/1sto22j/i_implemented_nag_normalized_attention_guidance/?
I feel like the v27 model is losing its anime style
The further away from V0.24, the less anime it is and the more 3D and 2.5D pollution in data there is :(
Your the stability king for a reason. You have experience on all preview models so I trust yours the most. The official turbo Lora is decent But it isn't as detailed and doesn't follow the prompt exactly unlike yours. Also this is the official NAG that's finally got anima support However, my experience with it has been a little bit mixed:
https://github.com/BigStationW/ComfyUI-NAG-Extended/tree/main
The official Turbo LoRa has the superior style preservation. RDBT the superior compositions. And something went horribly wrong with the NAG implementation. Hope it gets fixed.
@deitychaser I noticed that with I noticed that with NAG. To be honest, I don't think it's really great on anima I think it's more designed for distilled models like Flux Klein 9b/4b and z Imege turbo not really the base model with a distilled LORA.
@AnimaXx Maybe, not sure. It works better on non-distilled Chroma as a negative conditioning guidance than CFG though, in my experience. Simply because you can go up the NAG guidance to 6.0 or something which you can't do with CFG without frying your image, thus in my tests with NAG i could get more control.
Hi boss, is there any possibility that lora version will be released again? Being able to control the strength of lora/distilled model at any time is really useful.
For example, I also use the lore for distillation and it seems to be better, because the built-in lore in the model, it seems to me, has less power overall.
Probably, the reason I decided to release a ckpt because I think mixing distilled models is not a good idea. Distilled lora is not style lora. Distilled lora changes how model works.
@reakaakasky there seems to be many people thinking that RDBT makes @artist less effective, and actually it seems that you don't cite the artists in your work, and that's how your distill different from the official version, what do you think about this issue?
@reakaakasky so putting a distilled lora in a non-distilled base model or vice-versa, either is a bad idea?
@m4rbleye mixing different distilled models, e.g. 0.5 rdbt + 0.5 turbo
@jimzlf if I have 100x compute, I might consider taking care all 20k+ @ tags, but I don't.
Details
Files
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.