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。
RH体验地址（邀请码Invite Code: cptpovgo）：https://www.runninghub.ai/user-center/1993809672580788226?inviteCode=cptpovgo
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 💖!
Wanna leave me a tip? Drop by my discord channel for links and special perks. 想赞助一下？来我 Discord，里面有链接和特别福利哦！
Fanvue: https://www.fanvue.com/catlover1937
Ko-fi: ko-fi.com/catlover1937
Please do not redistribute without consent. 如非作者同意禁止搬运。
2026-07-10
✨ V4.0 Model Update
A short update note but lots of small stuff! This version includes a newly trained LoRA along with many quality improvements.
🔹 What's Changed
🧠 Integrated a newly trained LoRA for better overall generation quality.
✨ Reduced unwanted skin texture imperfections while preserving natural-looking fine details.
🚿 Reduced the chance of random body fluid artifacts appearing in NSFW generations.
🎨 Improved compatibility and support for external LoRAs.
🚀 For the best experience, pair this model with the new Krea2 Simple Workflow V3 or Minimal Workflow V3. Together they deliver noticeably cleaner outputs, better clarity, and improved overall image quality.
Fixed the FP8 conversion quality degrade issue
Fixed blurry wormlike background issue
✨ V4.0 模型更新
短简介大更新！本次版本加入了全新训练的 LoRA，并带来多项画质优化。
🔹 更新内容
🧠 整合了全新训练的 LoRA，进一步提升整体生成质量。
✨ 在保留丰富细节的同时，减少皮肤纹理瑕疵，让肤质更加自然干净。
🚿 降低 NSFW 场景中身体随机出现体液残留伪影的概率。
🎨 提升对外部 LoRA 的兼容性与表现。
🚀 强烈建议搭配全新的 Krea2 Simple Workflow V3 或 Minimal Workflow V3 使用，可显著提升画面清晰度、干净度以及整体生成质量。
修复FP8转换降画质问题
修复背景失真问题
2026-07-07
🚀 I also added the INT8 CONVROT version! In theory, it should deliver quality similar to Q8 GGUF while running at speeds close to INT8. ⚡
🚀 我新增了 INT8 CONVROT 版本！理论上，它能够提供接近 Q8 GGUF 的画质，同时拥有接近 INT8 的运行速度。⚡
好像很多人不知道每个版本代表着什么，It seems like many of you are unsure of what the models variants mean so here's my take for it:
I was able to rent a cloud server and done a special merge on BF16. Supporters can download version BF16 of this model directly from my discord link. DM me for discord permission tags. 我成功租到了一台云端服务器，并完成了这个模型的 BF16 特别合并版本。支持者现在可以通过我的 Discord 链接直接下载 BF16 版本。如需获取 Discord 下载权限，请私信我领取对应的权限组（Tag）。
FAQ:
Where is int8?
The mxfp8 is the int8, because of civitAI doesn’t have an int8 model type in the dropdown.
What are the special editions (models without Version tag)?
TLDR: download official editions (ones with version tags) for overall better experience.
The Special Edition are heavily inspired by my Asian girl character and style LoRAs. It was never originally intended for release—I created it purely for own enjoyment. However, after sharing some sample results in the community, quite a few people asked if I could release it, so I decided to publish it for those who are interested. It’s designed for a niche audience that enjoys the 网感 (internet aesthetic) beauty filter look.
Hope that clears things up!
2026-07-06
TDLR: V3 is so powerful that i was impressed when testing it. You can safely delete the previous versions now.
✨ Moody Krea2 V3 Release ✨
Hey everyone! 👋
V3 is here, bringing another solid upgrade over V2 with a stronger focus on realism while keeping the signature Moody look. 📸
🔥 What's Improved
🧑🎨 Increased overall realism and more natural-looking results.
🌄 Fixed the lazy background rendering issue from V2. Many of you reported the "worm-like" texture that sometimes appeared in blurred backgrounds—this has been significantly reduced.
💃 Improved human anatomy and more natural poses.
📝 Retrained several LoRAs using both Chinese 🇨🇳 and English 🇬🇧 captions, giving the model a better understanding of prompts and noticeably improving prompt adherence.
✨ Enhanced skin texture for results that look less AI-generated and more photographic.
⚙️ Workflow Update
I've also made a few adjustments to my recommended workflow to better support the model and get the most out of V3.
👉 Please give the new workflows a try for the best experience!
As always, thank you all for the feedback and bug reports—they've been incredibly helpful in making this release possible. ❤️
Enjoy, and have fun generating! 🚀
✨ Moody Krea2 V3 发布啦！✨
大家好～👋
V3 是基于 V2 的全面升级版本，在保留 Moody 风格的同时，进一步提升了整体真实感。📸
我只能说V3。。。很强。。。
🔥 本次更新内容
🧑🎨 整体画面更加真实、自然。
🌄 修复了 V2 中较常见的背景偷懒渲染问题。不少用户反馈虚化背景偶尔会出现类似「蚯蚓纹」的纹理，现在已经大幅改善。
💃 优化人体结构与姿势生成，让人物更加自然。
📝 重新训练了多组 LoRA，并同时使用 中文 🇨🇳 + 英文 🇬🇧 Caption，模型对 Prompt 的理解能力更好，Prompt Adherence 也有明显提升。
✨ 优化皮肤质感，减少 AI 感，让成图更加接近真实摄影效果。
⚙️ 工作流更新
我也针对模型调整了推荐工作流，让 V3 能发挥更好的效果。
👉 建议大家使用新的 Workflow，以获得最佳生成体验！
感谢大家一直以来的反馈和 Bug 回报，很多改进都是来自社区的建议。❤️
祝大家玩得开心，Happy Generating！🚀
2026-07-04
🎉🌸 Special Edition Drop / 特别版发布 💖
🇺🇸 English
Hello everyone! 👋
Here comes another Special Edition. Say no more. 😏
This release is trained on a heavily Xiaohongshu-inspired dataset, featuring East Asian women with beauty-filter aesthetics and curvier proportions. It's made specifically for a niche audience that enjoys this particular style. 💕
If you're not the target audience, feel free to skip this one and wait for the official V3, which is already in the works and will be the more balanced release for everyone. 🚀
Cheers, have fun, and enjoy! 🍻✨
🇨🇳 中文
大家好！👋
又来一个 特别版 啦，不多说了。😏
这次依旧是一个 小红书风格浓厚 的数据集，主打东亚女生、滤镜感颜值，以及更加丰满的身材表现。💕 这是专门为喜欢这种风格的小众用户准备的版本。
如果这不是你的菜，也完全没关系～可以直接等待正在制作中的 官方 V3，那会是更加均衡、适合大众使用的正式版本。🚀
祝大家玩得开心，玩得尽兴！🍻✨
2026-07-01
🎉🚀 Moody Krea2 V2 Official Release! 🌍✨
The wait is over! 💙 Moody Krea2 V2 is officially here!
The Oriental Edition was just a special edition released exclusively for my lovely Discord & Telegram community ❤️. It was made for a specific audience.
The Global Edition should be considered the official V2, featuring the most balanced finetuning I've achieved so far while preserving the signature Moody quality. 😎✨
🎮 GL HF, and enjoy! ❤️
⚠️ PS: About the MXFP8 version
The MXFP8 upload is my attempt at providing an INT8 version. Since Civitai doesn't currently have an INT8 model option, I selected MXFP8 instead.
I followed all the available conversion steps, but I wasn't able to fully verify that it works in every environment. As a result, I can't guarantee compatibility or functionality, so please use it at your own discretion.
🙏🎉🚀 Moody Krea2 V2 正式发布！🌍✨
终于来了！💙 Moody Krea2 V2 正式版 现已发布！
之前的 萌妹版版（Oriental Edition） 只是专门为 Discord 和 Telegram 群友准备的特别版本 ❤️，面向的是比较特定的受众。
而这次的 Global Edition 才是 真正的 V2 正式版 🌎，也是目前为止我调校得最均衡、最能体现 Moody 风格 的版本。😎✨
🎮 祝大家玩得开心，Enjoy！ ❤️
⚠️ PS：关于 MXFP8 版本
MXFP8 是我尝试制作的 INT8 版本。由于 Civitai 目前没有 INT8 的模型选项，所以只能选择 MXFP8 作为上传类型。
我已经按照目前已知的流程完成了转换，但 没有办法完整验证它是否能在所有环境下正常运行。因此，我无法保证兼容性或功能正常，如果有问题还请自行承担风险，感谢理解！🙏
2026-06-30
🌸 Krea2 V2 – Oriental Edition is Here! | Krea2 V2 东方版正式发布！ 🎉
🇺🇸 English
✨ Krea2 V2: Oriental Edition is here!
This SPECIAL EDITION was actually requested by members of our chat group. I originally made it just for my own entertainment and wasn't planning to release it, but after seeing how many people wanted it, I decided to share it with everyone. ❤️
This version focuses on beautiful East Asian female aesthetics with a strong Instagram and Xiaohongshu (RED) inspired vibe. 📸🌸
🔥 What's New
🤝 Huge kudos to @salttaro ! His Better Pussy & Poses V4 plays a major role in this edition on NSFW.
🎨 Added 2 newly trained Moody Aesthetic LoRAs, ported from ZImage for richer visuals.
🔞 Enhanced NSFW prompt adherence for better consistency and prompt accuracy.
NVFP4 Version.
🚀 A regular Krea2 V2 release is coming soon, so stay tuned!
Thank you everyone for your support! ❤️
🇨🇳 中文
✨ Krea2 V2 东方版 正式发布！
其实，这个版本是应聊天群里不少朋友的要求制作的。原本只是做给自己娱乐使用，并没有打算公开发布，不过看到大家都很期待，所以决定分享给大家。❤️
本版本主打 东亚女性美学，整体风格拥有浓厚的 Instagram 与 小红书（Xiaohongshu） 氛围，适合喜欢清新、时尚、高质感人像的用户。📸🌸
🔥 更新内容
🤝 特别感谢芋头哥 @salttaro ！本模型融合了 Better Pussy & Poses V4，对整体效果帮助非常大。
🎨 加入了 2 个全新训练的 Moody Aesthetic LoRA，并从 ZImage 移植整合，让画面氛围更加出色。
🔞 增强 NSFW Prompt 遵循能力，提高生成稳定性与提示词响应效果。
NVFP4 格式
🚀 普通版 Krea2 V2 也将在近期发布，敬请期待！
感谢大家一直以来的支持！❤️
2026-06-25
YES - IT SUPPORTS ALL RACES - I AM JUST LAZY BUILDING A MULTICULTURAL PROMPT DB.
RH体验地址：https://www.runninghub.ai/zh-cn/post/2070402594421436417/?inviteCode=cptpovgo
🚀 Moody Krea 2 Mix V1 is HERE! 🎉🔥
After lots of testing, it's finally time to release the very first version of Moody Krea 2 Mix! 💫
This is the initial release—I'm testing the waters with Krea 2, but I have to say... this base model is absolutely amazing. 🤯✨
I've already got several improvements and future training plans in the works, so this is just the beginning! Expect even better quality and refinements in upcoming versions. 💪🎨
🔓 NSFW is fully unlocked.
Have fun, experiment, and let me know what you create! Your feedback will help shape future updates. ❤️
Enjoy! 🚀🔥
🚀 Moody Krea 2 Mix V1 正式发布！ 🎉🔥
经过一段时间的测试，终于迎来了 Moody Krea 2 Mix 的首个版本！✨
这是一次初始版本，先放出来试试水。不过不得不说，Krea 2 的底模真的非常惊艳，潜力十足！🤯
后续我已经规划好了几轮新的训练和优化，所以这只是个开始，未来还会继续提升画质、细节和整体表现，敬请期待！💪🎨
🔓 NSFW 已完全解锁。
欢迎大家尽情体验、多多出图，也欢迎分享你的作品和反馈，你们的建议都会成为后续更新的重要参考！❤️
祝大家玩得开心！🚀🔥
PS: Workflow in this image - 图里有工作流 I will be posting an official workflow shortly. 我晚点发配套工作流。
Description
FAQ
Comments (45)
Anyone got frizzy hair problem with this model now??
The trick for creators who release multiple models is to pick one that works best and only upgrade when you get actual improvements.
int8在30系列显卡简直起飞,太快了
wtf new model every day. lol
Well not really new, just a new int8 format
not able to download the v3.0 model tried 3 times
"moodyKrea2Mix_v30.safetensorsFailed - Needs authorization"
well, you should report it to civitai. Alternatively you can download from my hg
I uploaded it without any problems, but sometimes the website works poorly or many features don't function correctly at all. Sometimes the speed is very slow. Civitai has been working very poorly for the past few years.
@catlover1937 unfortunately the bf16 version is not on your hf
@Ultrabad78527 my hg is running out of space, i thought BF16 is less wanted so
@catlover1937Hello! You can use this website to upload models "https://www.modelscope.ai/home" as far I know they provide unlimited cloud storage for models and datasets and also it is similar to hugging face.
@Wendy_Earth well i guess so... just got too many places to upload my models to.. RIP my bandwidth
For some reason i can't get fp8 version to work in forge-neo, I just get black squares. Can someone help me or can we get a fp16 version?
Try to run it with “--disable-sage”
@Vita I have ":: --disable-sage" in the webui-user.bat file on the line right before "call webui.bat" is that where its supposed to be?
@CaptainCrunch89 set COMMANDLINE_ARGS=--disable-sage
same as everyone use --disable-sage fp8 and even int8 can work in forge neo if you are updated even thought stability matrix
My ComfyUI version is 0.27.0. Using your V3.0 int8 convrot model, the image is very rough, while the V3.0 FP8 model has no problem. Does the int8 model require specific nodes?
No shouldn't be, it was fine in my testing though
@catlover1937 我也有同样的问题 v3 int8 convrot要比fp8糊的多，用其他模型包括krea2的int8 convrot版本都没有问题
@xinw9585 int8比FP8糊是一定的 FP8毕竟接近无损了
FormatApprox. QualitySpeedVRAMNotesBF16100.0%⭐⭐⭐🔴 HighestReference quality. No quantization loss.FP899.8–99.9%⭐⭐⭐⭐🟠 HighPractically indistinguishable from BF16. Best overall choice for RTX 40/50 series.INT8 (Native)99.6–99.8%⭐⭐⭐⭐⭐🟡 MediumVery small quality loss, excellent speed.INT8 ConvRot99.7–99.9%⭐⭐⭐⭐⭐🟡 MediumKeeps sensitive Conv/Rotary layers in higher precision, making it almost identical to FP8 while retaining INT8 speed.Q8 GGUF99.3–99.6%⭐⭐⭐⭐🟢 LowerExcellent compromise. Tiny quality drop.Q6_K GGUF98.5–99.0%⭐⭐⭐⭐🟢 LowSlight softening of fine textures and details.Q5_K GGUF97.5–98.5%⭐⭐⭐⭐🔵 Very LowDetails begin to noticeably degrade in complex scenes.Q4_K GGUF95–97%⭐⭐⭐⭐⭐🔵 LowestGood for experimentation, but quality loss becomes obvious.
There is too dark or old shape or color at her privates or nipples at my case. How can generate girl's privates to fresh young presents? :D
sometimes i get good one when i say young flawless pussy, smooth skin
@catlover1937 In my case, when my custom Lora draws a voluminous chest, the moody model always expresses the protrusion that comes into contact fabric area or covered nipples as very black, which is often a shame. I suggesting it should you make it to change it to a very bright pink color. That's my hope. :D Thanks for your kind good work.
Whenever i change the prompt, it generates a black screen. But if i clear all caches it works again. I'm confused..
(update): clearing caches doesn't work anymore.
are you using sage attention? or what is your workflow (may be some node problem), try using another workflow, use the workflow provided by catlover here.
@AI_Aurora I'm using moodyKrea2Simple_v20 workflow. Im looking in workflow overview and i searched for sageattention... SeedVR2 (Down)Load DiT ModelSeedVR2 Processes
attention_mode
sageattn_3
thats what came up.
@valeriexoxo sage attention 3 only works with 5000 series gpu. (if you send me the log of when you get the black output, then may be i can look into it)
@AI_Aurora Thank you i'm sorry to be a nuisance. So this is the black image logs
[32m[INFO][0m got prompt [32m[INFO][0m Model Krea2TEModel_ prepared for dynamic VRAM loading. 4999MB Staged. 0 patches attached. Force pre-loaded 249 weights: 627 KB. [32m[INFO][0m Model Krea2 prepared for dynamic VRAM loading. 12857MB Staged. 0 patches attached. Force pre-loaded 160 weights: 2824 KB. 0%| | 0/7 [00:00<?, ?it/s] 0%| | 0/7 [00:00<?, ?it/s, Model Initializing ... ] 0%| | 0/7 [00:00<?, ?it/s, Model Initialization complete! ] 14%|█▍ | 1/7 [00:00<00:01, 3.94it/s, Model Initialization complete! ] 14%|█▍ | 1/7 [00:00<00:01, 3.94it/s] 29%|██▊ | 2/7 [00:00<00:01, 4.41it/s] 43%|████▎ | 3/7 [00:00<00:00, 4.40it/s] 57%|█████▋ | 4/7 [00:00<00:00, 4.40it/s] 71%|███████▏ | 5/7 [00:01<00:00, 4.40it/s] 86%|████████▌ | 6/7 [00:01<00:00, 4.39it/s] 100%|██████████| 7/7 [00:01<00:00, 4.42it/s] 100%|██████████| 7/7 [00:01<00:00, 4.40it/s] [32m[INFO][0m Model WanVAE prepared for dynamic VRAM loading. 241MB Staged. 0 patches attached. Force pre-loaded 60 weights: 61 KB. /workspace/ComfyUI/nodes.py:1659: RuntimeWarning: invalid value encountered in cast img = Image.fromarray(np.clip(i, 0, 255).astype(np.uint8)) [32m[INFO][0m Model Krea2 prepared for dynamic VRAM loading. 12857MB Staged. 0 patches attached. Force pre-loaded 160 weights: 2824 KB. 0%| | 0/5 [00:00<?, ?it/s] 0%| | 0/5 [00:00<?, ?it/s, Model Initializing ... ] 0%| | 0/5 [00:00<?, ?it/s, Model Initialization complete! ] 20%|██ | 1/5 [00:00<00:03, 1.19it/s, Model Initialization complete! ][32m[INFO][0m Interrupting prompt 99b021dc-dfe3-4067-952f-5c84aa847279 20%|██ | 1/5 [00:00<00:03, 1.04it/s, Model Initialization complete! ] [32m[INFO][0m Processing interrupted [32m[INFO][0m [32mPrompt executed in 3.47 seconds[0m.
There is also a new problem. I've freed the model and node cache and unloaded models, changed the seed by one, and now i'm getting an image of loads of different colours. I'm also on RTX 5090
amazing detail, better body shapes, almost no distortions, and compatible with Loras !!!! LOVED IT!
keep the good work
@catlover1937 both your zit and krea2 models are top tier, and highly capable of generating sugary sweet aesthetics. thanks for sharing your amazing work.
Tested many models and this one has one of the best prompt adherence!
The only bias it has is with nudity even when you describe the outfit detailed, but that's more of a feature to some of us :')
Hi! Why when I use LORAS with your model it gives me lots of errors like those ones:
[WARNING] lora key not loaded: diffusion_model.layers.9.feed_forward.w1.lora_A.weight
[WARNING] lora key not loaded: diffusion_model.layers.9.feed_forward.w1.lora_B.weight
[WARNING] lora key not loaded: diffusion_model.layers.9.feed_forward.w2.lora_A.weight
[WARNING] lora key not loaded: diffusion_model.layers.9.feed_forward.w2.lora_B.weight
[WARNING] lora key not loaded: diffusion_model.layers.9.feed_forward.w3.lora_A.weight
[WARNING] lora key not loaded: diffusion_model.layers.9.feed_forward.w3.lora_B.weight
interesting, first time seeing this.
@catlover1937 yeah, it's annoying, only on some lora files, not all the ones
is this a turbo /distilled model , or a full step one ?
i think raw is only for lora training purposes, no one actually publishes full step raw models for Krea2. So it is a "turbo/distilled" model.
好像从BF16转换的Krea2 FP8版本有问题，难怪你们说的问题我好像没遇到过，看图：
Ok guys, I am not hallucinating. There's something wrong with the Krea FP8 converting from BF16. It's got huge loss in details, no wonder you guys are seeing things that I don't. Check Comparisons: https://civitai.red/posts/29693649
TLDR: FP8 SKIN LOOKS PLASTIC
Your work is fantastic; in my opinion, you are the only content creator who brings a unique "kawaii" touch, something no one else quite manages, when it comes to SFW and NSFW content.
I can only thank you for your effort, work, and dedication.
Catlover1937.... fantastic, magnificent, divine, these are the words that come to mind when I think of your name.💖
I was looking for a 'But'. So surprised I didn't find one. Cheers!