models/Hassaku XL (Illustrious) - v2.1fix

Hassaku XL (Illustrious) - v2.1fix

8/24/2025

3:23:09 AM

Related Keywords & Tags

anime,anime style stable diffusion,base model,characters,checkpoint,hassaku anime model,hassaku xl,hassaku xl (illustrious),hentai,ikena,illustrious,nsfw,sfw,v2.1fix,woman

Greyscale drawing of a confident female stormtrooper with red hair, black lips, and power armor in a dynamic pose with a detailed background.

A robot girl with intricate white and black mechanical joints and vivid red braided hair adorned with white flowers stands facing a black ornate framed portrait of a grumpy white cat with red ears and flower-decorated antler-like branches.

A dark-skinned woman with exceptionally long white hair and cat ears lounges elegantly on a deep-red velvet chaise, wearing a crimson top and glittering red skirt, holding a black cat in warm golden afternoon light.

Recommended Prompts

1girl, masterpiece, looking at viewer

Recommended Negative Prompts

signature

worst quality, bad quality

Recommended Parameters

samplers

Euler a

steps

cfg

resolution

1536x640, 1344x768, 1216x832, 1152x896, 1024x1024, 896x1152, 832x1216, 768x1344, 640x1536, 832x1224, 896x1152, 832x1216, 800x1248, 816x1232, 880x1232

vae

sdxl_vaeFix.safetensors

Recommended Hires (High-Resolution) Parameters

upscaler

R-ESRGAN 4x+ Anime6B

upscale

1.45 - 1.6

denoising strength

0.39 - 0.57

Tips

The model is trained with minimal disruptive elements; if such appear, use 'signature' as negative prompt.

Exclude metadata and franchise tags such as 'highres' and 're:zero kara hajimeru isekai seikatsu'.

Use prompt order with number of persons first, then character names, followed by other attributes for optimal results.

NoobAIs Loras outperform Illustrious loras when used with this model.

Version Highlights

Because my 2.2 Version needs a bit more time then expected. I made this fixed version of 2.1. Issues of this version was a to high style merge overall, resulting in unexpected issues with samplers and unwanted side effects.

I reduced the style merge significatly and merged style A in, because style A is overall more liked. Not only because its more stable, but also for its simpler but still detailed style.

2.1fix should have a lot less issues sampler side and resembles style A more or less in some aspects. Drawback of mergin A back is, that newer characters are getting erased in this fixed version. But that seems for most not that big of a problem, especially with the use of loras.

Note: Fixed Style is weaker on the fix version, the variety of different styles that can appear is much higher. If you don't like that faces can appear more realistic, put nose into the negatives.

Creator Sponsors

You can run Hassaku XL and use its API on SinkIn: https://sinkin.ai/m/JWknjgr

My Discord for everything related to anime models and art.
You can support me on my patreon.

Hassaku aims to be a anime model with a bright and distinct anime style.

You can run Hassaku XL and use its API on SinkIn: https://sinkin.ai/m/JWknjgr

My Discord for everything related to anime models and art.
You can support me on my patreon.

____________________________________________________________
Supporters:

Thanks to my supporters Riyu, SETI and Kodokuna on my patreon!

____________________________________________________________
Using the model:

The model is trained using images with minimal disruptive elements such as floating text, logos, speech bubbles, and signatures. If any of these elements are present in an image, please include the prompt "signature" as a negative prompt
Metadata and franchise tags are excluded—please do not use them. Tags such as "highres" and franchise-related tags like "re:zero kara hajimeru isekai seikatsu" are not used in training
Look at the example images, to see how the model should be used
Model is trained to be not complicated in use, tag order:
First are the number of person, then character names, rest.
Example: 1girl, rem \(re:zero\), standing, masterpiece, upper body
NoobAIs Loras working for the most part better in compare to Illustrious loras
Here are some resolution options for SDXL:
- 1536 x 640
- 1344 x 768
- 1216 x 832
- 1152 x 896
- 1024 x 1024
- 896 x 1152
- 832 x 1216 (most recommended)
- 768 x 1344
- 640 x 1536

______________________________________________________

Version and License info:

Below V1 merges ANIMAGINE XL 3.0
Version V1 use Illustrious-XL & WAI-NSFW-illustrious-SDXL with additional training
V2 is trained on its own and don't include any extra merge. Base was V1, it is trained to include newer or missing characters. It was also used for further training tests
V3 is a merge of Illustrious-XL and V2, to fix some issues and was also trained to include newer or missing characters
All Models using the Fair AI Public License 1.0-SD license.