Introducing WAI-illustrious-SDXL-v16! A Deep Dive into Civitai Licenses

clear_an animated character with silver hair and blue eyes wears a red jacket with a badge, standing in front of a mountainous landscape at sunset or sunris_1
  • WAI-illustrious-SDXL-v16 is the latest anime model.
  • Natural evolution in color and detail.
  • Wonderful, supple skin/hair texture.

Introduction

Hello everyone, I’m Easygoing.

Today, I’m excited to introduce WAI-illustrious-SDXL-v16, a highly popular anime model for SDXL.

clear_an animated female character with silver hair and blue eyes wears a red jacket with a white fur collar, set against a backdrop of a sunset or sunrise 00001
waiillustrioussdxl_v160_clear

What is WAI-illustrious-SDXL-v16?

The WAI-illustrious-SDXL series, published by WAI0731, is arguably the most popular anime model currently available for image generation.

gantt
    title WAI-illustrious-SDXL series
    dateFormat YYYY-MM-DD
    tickInterval 6month


	v1: 2024-10-06, 2025-12-27
	v2: 2024-10-23, 2025-12-27
	v3: 2024-10-28, 2025-12-27
	v4: 2024-11-02, 2025-12-27
	v5: 2024-11-04, 2025-12-27
	v6: 2024-11-10, 2025-12-27
	v7: 2024-11-17, 2025-12-27
	v8: 2024-12-20, 2025-12-27
	v9: 2025-01-15, 2025-12-27
	v10: 2025-02-04, 2025-12-27
	v11: 2025-02-14, 2025-12-27
	v12: 2025-03-05, 2025-12-27
	v13: 2025-04-03, 2025-12-27
	v14: 2025-05-08, 2025-12-27
	v15: 2025-09-01, 2025-12-27
	v16: crit, 2025-12-18, 2025-12-27

While the series undergoes frequent updates, the latest version, WAI-illustrious-SDXL-v16, was released on December 18, 2025.

V16 has been trained on the latest characters and features significantly more accurate depictions of eyes and hands, making it a flawless, all-purpose anime model.

Comparing v16 vs. v15

Let’s compare the output of the new v16 model against the previous v15.

In the images below, the left side is the newly released WAI-illustrious-SDXL-v16, and the right side is the previous v15.

Mountain Sunrise

Comparison of mountain sunrise illustrations: v16 Comparison of mountain sunrise illustrations: v15
WAI-illustrious-SDXL-v16 | WAI-illustrious-SDXL-v15

Present

v16 vs v15 comparison: v16 shows brighter faces and clearer expressions v15

Close-up

Close-up of hair: v16 (left) shows finer, more organized lines

Comparing the two, v16 produces cleaner overall output with brighter facial lighting.

When looking at the hair close-up, you can see that v16 features sharper, more refined lines and a more natural flow.

Derivative models of the Illustrious-XL series often end up with a "hard, plastic-like" texture when trying to increase quality. However, the WAI-illustrious-SDXL-v16 model manages to capture a soft, supple, and elastic texture.

Improving WAI-Illustrious-SDXL-v16!

While WAI-Illustrious-SDXL-v16 is already highly polished, I attempted to push the expression even further with some adjustments.

The left side is my adjusted WAI-illustrious-SDXL-v16_clear model, and the right side is the original v16.

Comparison and difference map between the custom

In this "clear" version, besides subtle model tuning, I upgraded the VAE to an FP32 format specifically designed for anime, which I believe further improves detail and color depth.

SDXL Anime-Specific VAE:

Image Difference Tool:

What is the License for WAI-illustrious-SDXL?

Now, let's look at the licensing for the WAI-illustrious-SDXL model.

It is currently available on Civitai, TensorArt, and Seaart AI.

Civitai

Civitai license display for WAI-v16 showing

Tensor Art

Terms of use on TensorArt showing commercial use allowed but generation services prohibited

Seaart AI

Screenshot of license details on Seaart AI

While each platform uses its own specific UI for licensing, they can be summarized as follows:

  • Model Merging: Allowed
  • Commercial Use of Outputs: Allowed
  • Use in Generation Services (SaaS): Prohibited
  • Selling the Model or Derivatives: Prohibited

Beware of Civitai’s License Labels!

The most important thing to watch out for is the license notation on Civitai.

Civitai labels it as License: Illustrious License, but in reality, no such single license exists. The Illustrious-XL series actually uses different licenses for different versions:

Reference: Illustrious-XL Series Licenses

The link for the "Illustrious License" on Civitai currently points to the Fair AI Public License 1.0-SD. This refers only to v0.1 and has likely not been updated properly since v1.0 and later versions were released.

License Settings are at the Creator’s Discretion

While licensing is ultimately determined by the creator, Civitai has a system where selecting a base model type automatically applies the original model's license, which the uploader cannot always manually edit.

Civitai page for Animagine-XL_3.1 showing CreativeML Open RAIL++-M
Animagine-XL_3.1 License Display
Civitai page for Animagine-XL_3.1 showing Fair AI Public License 1.0-SD in the description
The description says one thing, the UI says another. The description is likely the correct one.

If a creator wants to set a license different from the auto-applied one, their only choice is to write the correct license in the model description.

Therefore, since Civitai’s license labels are frequently inaccurate, always read the model description carefully. If you’re still unsure, cross-check the model on other hosting sites.

WAI-illustrious-SDXL-v16_clear Workflow

Because of these licensing complexities, I won't be publicly distributing the WAI-illustrious-SDXL-v16_clear model file. Instead, I’m sharing the merge recipe for it.

waiillustrioussdxl_v160_merge_20251227

Required Models:

Using this workflow, you can generate images with the WAI-illustrious-SDXL-v16_clear settings. If you like the results, you can also save the model locally by connecting a Save Checkpoint node.

ComfyUI Save Checkpoint node

The official WAI-illustrious-SDXL-v16 guide recommends using a short Negative Prompt, but since I found it made very little difference, I have omitted it in this workflow.

If you want to use a Negative Prompt, please reconnect the ConditioningZeroOut node accordingly.

Benefits of Omitting Negative Prompts:

High-Quality Workflow with Upscaler & Detailer

Finally, here is a workflow that goes beyond standard generation by incorporating Upscaler and Detailer processes to produce high-definition illustrations.

ComfyUI workflow combining Upscaler, Detailer, and SAM3

Required Models:

This is essentially the same as the SAM3 workflow I introduced previously, but I’ve updated the SAM3 processing to use the comfyui_sam3 custom node to optimize VRAM usage. In my environment, this custom node is much less prone to "Out of Memory" errors.

Please note that you still need to download the SAM3 model separately; refer to my previous article for instructions on how to do that.

Summary: Give WAI-illustrious-SDXL-v16 a Try!

  • WAI-illustrious-SDXL-v16 is the latest anime model.
  • Natural evolution in color and detail.
  • Wonderful, supple skin/hair texture.

Today we looked at WAI-illustrious-SDXL-v16. It’s a definitive model for SDXL—if you're ever unsure which model to use for anime illustrations, you really can't go wrong with this one.

Anime illustration showing soft skin and hair textures via WAI-v16
Look at that soft, supple texture!

While new generation models like Z-Image are starting to emerge, the overwhelming polish of WAI-illustrious-SDXL-v16 suggests that SDXL’s reign might continue for a while longer.

Thanks for reading!