v-pred_v2.0
Carefully Curated Danbooru Dataset (up to September 2025)
Base Model: NoobAI-XL (NAI-XL) V-pred-1.0-Version
When necessary, please use Rescale CFG with values up to around 0.7 (especially for NSFW content).
Quality Tags:
very aesthetic > masterpiece > best quality > good quality > normal quality > worst quality
If you feel like it, you can support me on Ko-fi. Thank you.
Online Generator:https://www.seaart.ai/ja/models/detail/a106ce42854f7b7f65523ad10e03caa8
v-pred_v1.1
Dataset:50k AI-generated images from Aibooru and Pixiv + 10k images from Danbooru
Prompts
people, characters, copyright, style, general tags, rating, masterpiece, best quality, good quality, newest
Negative Prompts
lowres, worst quality, bad quality, bad anatomy, sketch, jpeg artifacts, signature, watermark, old, oldest
Base model:NoobAI-XL (NAI-XL) V-pred-1.0-Version + Illustrious XL 1.0
v-pred_v1.0
Base model:NoobAI-XL (NAI-XL) V-pred-1.0-Version
Use v-prediction and Zero Terminal SNR.
For information on v-prediction and how to use it in ComfyUI, refer to the NOOBAI XL Quick Guide by L_A_X.
As of November 20, checkpoints supporting v-prediction are not available for use in stable-diffusion-webui.
Outside of ComfyUI, you can use the webui's dev branch (likely requiring a .yaml file for each checkpoint) or Forge and reForge.
Additionally, make sure to select "Zero Terminal SNR" under Settings → Sampler parameters → Noise schedule for sampling.
v3.1
Base model:NoobAI-XL (NAI-XL) Epsilon-pred 1.1-Version
v-pred_v0.1
Base model:NoobAI-XL (NAI-XL) V-pred-0.6-Version
v3.0
Base model:NoobAI-XL (NAI-XL) Epsilon-pred 1.0-Version
v2.1
Base model:NoobAI-XL(noobaiXLNAIXL_epsilonPred075 x 0.7 + noobaiXLNAIXL_epsilonPred05Version x 0.3)
v2.0
Base model:NoobAI-XL (NAI-XL) Epsilon-pred 0.5-Version
v1.0
Base model:Illustrious-XL-v0.1
After training, this model merged Lucereon into the U-net.
(0,0.3,0.25,0.2,0.15,0.1,0.15,0.2,0.25,0.3,0.3,0,0.05,0.15,0.25,0.3,0.25,0.15,0.05,0)
License
Description
FAQ
Comments (15)
Best Illustrious-XL model for me so far, good balance between variety and quality stability. The most impressive thing is when you use artist styles, this model won't have uncontrolled watermarks and backgound texts appear as frequently as other illustrious models. Great job.
This is a fantastic model, very prompt adherent, great style. Weirdly it is SDXL but it seems to work better with PONY lora then SDXL. Are there some details about why in the base Illustrious-XL?
you are saying pony loras work with this?
@PM786 Yep, sometimes not as exactly as it does when using a pony model but it does have a general impact which can apply specific concepts or characters. You might need to increase the strength of the lora to 1.2 sometimes but 1 usually works.
@Evivrus thank you so much <3 and which illustrious model is best for you?
@PM786 So far I have been enjoying Obsession by @rqdwdw but MorgyMix-illustriousXL and illustrious-flying both have good base styles.
What's the recommended VAE? No matter which one I choose, I keep getting these weird green patterns embedded in the images.
so the problem is not in the VAE
My new favorite!
I tried every single illustrious model on civitai and this is the best one.
7 days ago
does it stand the test of time with all theses bieautiful merges now ?
I always thought you should never use ai in training data, doesn't model repeat same mistakes when its trained on ai data?
it depends if they cherrypicked ai images that doesn't have too many mistakes.
@AkioAI I always feel that using AI-generated images to feed into AI affects its aesthetic scoring, because even if the image quality is good, SD still struggles to generate beautiful perspective shots, battle scenes, scenery, and other high-aesthetic-rated pictures,This leads to Stable Diffusion being able to generate 'beautiful characters' but unable to truly create a harmonious aesthetic connection between the characters and the background, scenes, or objects.
@Clipbrd True. I know a lot of AI images get retouched and etc to have really good perspective to the point you can't even tell it is AI. I do agree that it loses value within the model that is why I think it should be mix, AI images and images from Gelbooru or Danbooru.
Details
Available On (14 platforms)
Same model published on other platforms. May have additional downloads or version variants.