I just made a new Telegram Group, drop in for chats, questions, feedbacks or just to share your work. I will be sharing prompts and pics there as CIVITAI has been removing my work repeatedly. 要词加群。
Coffee?: Ko-fi.com/catlover1937
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