I just made a new Telegram Group and Discord group drop in for chats, questions, feedbacks or just to share your work. 欢迎加入纸飞机和Discord群组。

喜欢我的作品吗？🎨 我会持续免费分享更多优质内容✨，欢迎支持一下，请我喝杯咖啡吧！💖 Like my work? 🎨 I share all my work for free ✨, so support me by buying me a coffee 💖!

Ko-fi: ko-fi.com/catlover1937

Fanvue: https://www.fanvue.com/catlover1937

Please do not redistribute without consent. 如非作者同意禁止搬运。

2026-05-06

HF 抱脸: https://huggingface.co/catlover1937/moody-real-v6

✨ Moody Real V6 Release ✨

Hey everyone 👋
Super excited to bring you my newest release — Moody Real V6 🎉

This version comes after a lot of refinement and experimentation. The goal was simple:
👉 Push realism further while introducing a subtle touch of fantasy that previous versions didn’t quite capture — without losing the signature Moody look you expect.

🔧 What’s Improved?

Over this update, I went back and revisited every LoRA used from V1 → V5:

Rebalanced and adjusted LoRA weights ⚖️
Retrained stylistic LoRAs with a cleaner, higher-quality dataset 🧠
Retrained body anatomy LoRA with a selective dataset ⬆️
Optimized overall LoRA integration to better support character LoRAs 🎭

🚀 What’s New?

💡 Lighting / Visuals / Colors
- More refined lighting
- Richer colors (less washed-out)
- Stronger overall visual impact
😊 Faces & Expressions
- Improved facial structure
- Better makeup rendering
- More natural and expressive faces
🔗 LoRA Compatibility
- Much better support for character LoRAs
- Better LoRAs compatibility support
🚫 NSFW Weight Adjustments (Again)
- Further reduced NSFW bias
- Retrained body anatomy LoRA with a more thoughtful dataset
- You can now safely use terms like “large breasts” without forcing NSFW outputs
- Still retains the ability to generate artistic nudity 🎨

🛑 Quick Note

Please stop saying “character LoRAs don’t work” and maliciously attacking me ❌ 😭

If you check my gallery, you’ll clearly see consistent characters across images —
👉 That’s proof they do work when used correctly.

⚠️ IMPORTANT — LoRA Usage Guide (PLEASE READ 🙏)

I’ve said this many times (probably 9000+ 😅), but let’s make it clear again:

✅ Recommended: Use ZIB LoRAs
→ Works perfectly up to 1.25 weight
⚠️ If using ZIT LoRAs:
→ Keep weight BELOW 0.4
🔧 Still not getting correct likeness?
→ Use my face detailer node in the workflow
→ Inpaint the face with your LoRA for better accuracy

❤️ Final Words

Your support truly keeps me going 🙌
Really appreciate everyone who’s been following the Moody series.

Have fun generating…
Moody gals are waiting 😏✨

✨ Moody Real V6 发布 ✨

大家好 👋
很高兴带来我的最新版本 —— Moody Real V6 🎉

这次更新投入了大量时间进行优化与调整。目标很简单：
👉 在保持系列一贯超真实风格的基础上，加入一点点之前版本没有的轻微幻想感。

🔧 本次优化内容

我重新检查并调整了 V1 到 V5 所有使用过的 LoRA：

重新调整权重 ⚖️
使用更干净的数据集重新训练风格 LoRA 🧠
选择性重组训练集，重新训练人体结构LoRA ⬆️
优化整体 LoRA 结构，提升角色 LoRA 兼容性 🎭

🚀 更新亮点

💡 光影 / 画面 / 色彩
- 光影更加自然细腻
- 颜色更丰富（不再发灰）
- 整体视觉效果更强
😊 人脸与表情
🔗 LoRA 兼容性提升
- 更好支持角色 LoRA
- 使用LoRA更不容易崩图
🚫 NSFW 权重再次降低
- 进一步减少 NSFW 倾向
- 重新训练人体结构 LoRA（更合理的数据集）
- 现在可以放心使用 “大胸” 等词，而不会直接变成 NSFW
- 同时仍然保留艺术裸体表现能力 🎨

🛑 说明一下

请不要再说 “角色 LoRA 不工作” ❌ 并且恶意攻击我 😭

我的作品展示里已经证明：
👉 只要使用正确，是完全可以稳定出同一角色的。

⚠️ 重要 — LoRA 使用说明（请认真看 🙏）

这个我真的说过很多次了（大概9000次😅）：

✅ 推荐使用 ZIB LoRA
→ 权重最高可到 1.25
⚠️ 如果使用 ZIT LoRA
→ 权重请 低于 0.4
🔧 如果角色不像？
→ 使用我的 workflow 里的 face detailer 节点
→ 用你的 LoRA 进行脸部重绘

❤️ 最后

感谢大家一直以来的支持 🙌

祝大家玩的开心，生成愉快——
Moody 女孩等你来生成 😏✨

2026-04-02

🌟 What's New in V5?

In V3 and V4, I focused on improving the model based on V2 while trying to preserve the artistic elements and unique styles that Moody Real maintained in V1 and V2 🎨. Despite my efforts, I wasn’t fully satisfied with the results 😅.

In V5, I’ve retrained several LoRAs with a better dataset and refined tagging 🗂️, while also reducing the weight of NSFW content 🔞 to better preserve the artistic qualities ✨. As I’ve mentioned before, this model is designed for softcore generation with light NSFW capability.

Finding the perfect balance between quality, style, and NSFW sensitivity has been a challenge ⚖️, but I believe V5 achieves a much closer equilibrium ✅. I hope you enjoy this version, and as always, your support is greatly appreciated 💌!

🌟 V5 新特性

在 V3 和 V4 中，我尝试在 V2 的基础上提升模型质量，同时保留 Moody Real 在 V1 和 V2 中的艺术元素和独特风格 🎨。尽管努力了，我对结果还是不太满意 😅。

在 V5 中，我重新训练了几个 LoRA，使用了更好的数据集和精细标签 🗂️，同时降低了 NSFW 内容的权重 🔞，以更好地保留艺术感 ✨。正如我之前提到的，这个模型主要用于软核生成，并带有轻量 NSFW 能力。

在质量、风格和 NSFW 敏感度之间找到完美平衡一直是个挑战 ⚖️，但我相信 V5 在这方面做得更好了 ✅。希望你喜欢这个版本，非常感谢大家一直以来的支持 💌！

GGUFs:

https://huggingface.co/Gthalmie1/moody-real-mix-v5-dpo-gguf

please show your thanks directly @CH11442 !

2026-03-08

gguf versions of moody-real-mix-v4 now: https://huggingface.co/Gthalmie1/moody-real-mix-v4-dpo-gguf please show your thanks directly @CH11442 !

Happy women's day!

Hello, HEADS UP, Moody REAL V4 just dropped!

What’s new in V4?

A HUGE improvement on image quality due to better trained LoRas.

PLUS

In Real V4, I’ve baked in ZImage DPO, which enhances color vibrancy and lighting — areas where ZImage Turbo base model falls short, especially when running heavily loaded LoRAs, a lot of the colors are lost. If you’re interested in the technical approach behind it, I’ve included the reference link. I think this is a great addition that makes the outputs more eye-catching while still maintaining a natural look.

If you’d like even more vibrancy, you can also download the DPO LoRA and stack it on top of Real V4. Just make sure to experiment carefully and know when to stop — don’t go too crazy haha.

On top of the DPO integration, I’ve made a few minor image quality adjustments. I specifically tried to reduce the strange “liquid” effect on skin textures, which I personally always found annoying. It’s difficult to fully eliminate due to ZImage Turbo’s low-step nature, as fine details are harder to resolve, but I’ve reduced it as much as possible. As a result, you may notice the skin appears slightly more matte than before, giving it a more realistic, real-life texture.

Real V4 is compatible with the ZIT Simple workflow as well as the ZIB + ZIT workflow . A image to image ZIT workflow is in the work.

As always, stay safe, boys and girls. And your support is greatly appreciated ❤️

大家好，注意啦！REAL V4 正式发布了。

V4 有什么新内容？

大大增强了画质，得益于我用更好的数据练的LoRas。

在 V4 中，我内置了ZImage DPO，它增强了色彩饱和度和光影表现——这正是 ZImage Turbo 原本的诟病，特别是在高负载 LoRA 情况下比较明显的色彩丢失的问题。如果你对技术实现方式感兴趣，我也附上了相关参考链接。我认为这是一次非常不错的融合，让画面在更吸睛的同时依然保持自然感。

如果你希望画面更加鲜艳，也可以额外下载 DPO LoRA 叠加在 V4 上使用。不过一定要自己多做实验，掌握好“适可而止”的程度，别叠得太猛哈哈。

除了 DPO 之外，我还对画质做了一些小调整。我尝试减少皮肤上那种奇怪的“液化感”纹理，这个问题我一直觉得很烦人。由于 ZImage Turbo 本身是低步数模型，细节收尾比较难彻底完成，所以很难完全消除，但我已经尽可能去优化。因此你可能会感觉皮肤质感比以前稍微更偏哑光一些，整体更接近真实生活中的皮肤效果。

和之前一样，Real V4 兼容 ZIT Simple 工作流以及 ZIB + ZIT workflow 工作流。

一如既往，大家注意安全。非常感谢你们一直以来的支持 ❤️

2026-02-24

GGUF Verions Provided by @CH11442 , please show your thanks directly @CH11442 !

Hello everyone,

I know Moody Real Model has needed some attention for a while and was temporarily deprioritized due to other releases, but here we are — V3 is finally here. With V3, you can expect all the updated ZiB LoRAs baked in, similar to Moody Porn Mix V9, which means a high level of LoRA compatibility if you enjoyed the results from V9.

Highly recommended to use with my ZIB + ZIT workflow

FAQs:

Why does Moody Real exist? What’s its purpose? Let me restate the original goal of this model:

1) It focuses on the stylistic elements of generating beautiful characters.

2) It is capable of producing artistic nude (softcore) images with solid body anatomy. If you are looking for explicit hardcore content, this is not the right model — please use my other model for that.

3) Improved LoRA compatibility.

Why are your demo images mostly Asian-oriented? Can this model only generate Asian characters?

No — Moody Real can generate people of all races. I’ve explained this a few times already. The reason most examples lean Asian is simply because I’m less confident in representing Western beauty standards accurately and prefer not to misrepresent them. If you generate something stunning, especially showcasing different aesthetics, feel free to share it. I’d be happy to pin outstanding results to highlight what the model can do.

Is this compatible with the ZIB + ZIT workflow?

Yes — in fact, it is highly recommended to use this model with ZIB + ZIT if you are not creating explicit hardcore content instead of using Moody Porn.

大家好，

我知道 Moody Real Model 已经有一段时间没有更新了，因为之前优先在做其他版本，不过现在 V3 终于来了。V3 内置了所有最新更新的 ZiB LoRAs，和 Moody Porn Mix V9 类似，如果你喜欢 V9 的效果，那么在 LoRA 兼容性方面你也可以期待同样高水平的表现。

极力推荐与我的 ZIB + ZIT 工作流一起使用。

常见问题：

为什么会有 Moody Real？它的定位是什么？

我再重新说明一下这个模型最初的目标：

1）专注于生成漂亮角色的风格表现。

2）可以生成具有良好身体结构的艺术裸（偏软核）作品。如果你想要的是非常露骨的硬核内容，那这个模型并不适合，请使用我其他专门的模型。

3）更好的 LoRA 兼容性。

为什么你的示例图大多偏亚洲风格？这个模型只能画亚洲人吗？

不是的——Moody Real 可以生成所有种族的人物。我已经解释过很多次了。之所以示例图偏亚洲风格，只是因为我对西方审美的把握没有那么自信，不想做得不准确。如果你生成了非常出色的作品，尤其是不同审美风格的，欢迎分享，我也很乐意把优秀作品置顶展示。

它兼容 ZIB + ZIT 工作流吗？

是的，事实上如果你不是制作硬核内容，我强烈建议在 ZIB + ZIT 工作流中使用这个模型，而不是使用 Moody Porn。

2026-02-04

ZIT V2

Hello AI Artists! I know I’ve owed you all a long-overdue update on this model, so here it is. This new release now includes some of my later LoRAs that were already baked in my other model.

To clarify, this model produces output that’s much closer to the ZIT base model, with added enhancements for characters and body parts, and it places a stronger emphasis on artistic quality rather than XXX elements. If you’re specifically looking for XXX content, feel free to check out my other model. 🙂

你好，AI 艺术家们！我知道我欠大家一个迟来的Moody Real模型更新，所以现在它来了。这次的新版本加入了一些我后期制作的 LoRA，这些内容之前已经整合在我的另一个 model中。

简单说明一下：这个模型的输出效果更接近 ZIT 基础模型，同时在角色和身体部位方面进行了增强，并且更加侧重于画面的艺术性，而不是 XXX 元素。如果你主要在寻找 XXX 内容，欢迎查看我的另一个 model。🙂

2026-01-01:

ZIT V1

Happy New Year!

This is the SFW counterpart to my other model—essentially everything from that model except MysticXXX , as it tends to have a very strong effect on styles (?). The goal is to offer a more compatible realistic base model that works well with character LoRAs or any other LoRAs, while reducing the risk of image distortion.

All display photos are generated without detailers/inpaint, and with workflow and prompts embeded.

PS: It still performs better than the turbo base model when it comes to nudes :).

That’s it—happy generating! Feel free to share your results below; I’d love to see how others are using this model.

ZIT V1

新年快乐！

这是我另一个模型的 SFW（安全向） 版本——基本包含了那个模型的所有内容，除了 MysticXXX，因为它对风格的影响过于强烈（？）。本模型的目标是提供一个兼容性更高的写实基础模型，能够更好地配合角色 LoRA 或其他 LoRA 使用，同时降低图像失真的风险。

所有展示图片均未使用 detailer / inpaint，并且已嵌入完整的工作流和提示词。

附注：在裸露表现方面，它的效果依然比 turbo 基础模型更好 🙂。

就这些啦——祝生成愉快！欢迎在下方分享你的作品，我也很期待看看大家是如何使用这个模型的。

My very first model, making it for fun.

Guidance: 7-12

Steps: 25 - 35

Sampler: DPM++ 2M

Hi-res fix 2x recommended. Adetailer on face was used.

Enjoy.

v5. Re-done the model to be more flexible for loras.

v4. Emphasize on drawing detailed hands. Change to a more mature face model so the photos don't get detected as under-aged.

v3: minor improvements (hands and face)

v2: skipped because too unstable

v1: first release

Files with SHA256: 618ce2db8f6819505c854f4ce44d28a3ca752e4464a55606c8f0c3df8906d761