(2026-04-19) Turbo V3.0 sometimes produces images with fine patterns, textures, or a slightly blurry/soft look. After testing and tweaking, this version has clearer image quality and better sense of depth and space compared to the previous one. NSFW performance remains at the same level.
In 3.0, the generated images have fine mesh-like stripes and a warmer color tone. However, these can be preserved after downloading. For complex XXOO actions/poses, the probability of the image breaking down or deforming is theoretically lower.
(2026-04-19)Turbo V3.0 生成图片，有些图片会有图案细纹和模糊感。经过我的测试和调试，这个版本会比之前的版本画质更清晰，更有空间层次感。NSFW方面保持相同水平。
3.0生成图片有细小的网状条纹，再就是色调偏暖。但是下载了可以保留下来，XXOO复杂动作，出图崩坏概率理论上会更低。
(2026-04-19) V3 Turbo is finally here after more than a month of training!
When I originally released the V3 Base, I actually wanted to release the Turbo version at the same time. However, the LoRA dataset was simply too large and might not have been compatible with the Turbo model.
The Base version still has the best image quality, but it’s quite slow to generate. Some people suggested using acceleration LoRAs, but I haven’t tried that myself.
If you’ve already downloaded the Easy Prompt pack from before, you can keep using your old prompts — they work exactly the same, no need to download it again.
I’ll update and add more AI example images whenever I have time.
I've also released Turbo V3 LoRA version, which allows for free control of NSFW weight in conjunction with the original Turbo model. The weight of this main model is 0.75.
Easy prompt for V3 download
Eason Zimageturbo workflow download
Eason ZimageTurbo realistic NSFW｜淫模 V3 LoRA download
My setting:
（2026-04-19）v3 turbo经过一个多月的训练终于来了，原本发布V3 BASE的时候我是希望能同时发布Turbo，由于Lora的素材量太大可能无法被Turbo 大模型所兼容。Base版本的模型画质是最好的，但是出图速度慢，有人提倡可以使用加速LORA，但是我没有尝试过。之前下载了傻瓜提示词（Easy prompt）的朋友可以继续使用原有的提示词，是一样的，不必重复下载。我会有空更新一些AI例子图片。
我也释放了Turbo V3 LoRA版本，可以配合原始Turbo模型自由的控制NSFW强度，本大模型强度为0.75
傻瓜提示词 V3 download
Eason Zimageturbo workflow download
Eason ZimageTurbo realistic NSFW｜淫模 V3 LoRA download
我的设置：
Easy promt for V3 download
Eason ZimageBASE workflow download
(3-7-2026) After nearly 5 weeks of hardcore, non-stop training, V3 is finally here! This beast clocked in at a mind-blowing 2.36 million steps — yeah, you read that right.
Compared to V2, which had all sorts of sexy outfits, lingerie, and heavy NSFW gear, V3 strips all that away. No more clothes cluttering things up. I also ditched every single AI-generated image from the training data to push for maximum realism. The focus is purely on authentic NSFW performance: genitals look way more lifelike and detailed, double-person sex scenes/positions feel stronger and more dynamic, and overall anatomy + motion just hits different.
Of course, with super complex poses or crazy action, it can still break sometimes , but honestly? After tons of testing, I'm super happy with V3. It feels like it has huge potential — the NSFW output is insanely real, rich, and intense. Just check out the example images... you'll feel the raw power right away.
Download it, fire up some generations, and enjoy the ride! 🔥 Let me know what crazy stuff you create~
I also release V3 Lora Version, Important note: This LoRA is not compatible with Z-Image Turbo — throwing it on Turbo will usually result in total snowflake/noise artifacts or just broken garbage. Don't waste your time trying unless you're into abstract chaos 😂.
It's built specifically for Z-Image Base, where you can freely crank the LoRA strength anywhere from 0.7 to 1 (or even higher) for super flexible control over style, details, and that raw NSFW intensity.
https://civarchive.com/models/2359268
My setting:
(3-7-2026) 经过将近5个星期的超负荷训练，V3版本终于发布了，训练步数达惊人的236万步。它与V2不同的是，V2有各种情趣衣服和NSFW的内容，而V3则没有那些衣服，同时为了更真实移除了所有AI图的训练素材，更专注于NSFW方面的训练，比如阴部显得更加真实，双人做爱运动也有所加强，当然也有时候会因为动作复杂而崩图。但是经过我的测试，我对V3非常满意，我感觉它是非常有潜力的模型，因为她的NSFW的表现非常真实和丰富，当你看见例图你也能感受的到NSFW的强度。欢迎你下载和生成图片，enjoy it！
同时，我也释放出炼V3的BASE Lora版本，注意，这个LORA与Turbo 模型并不兼容，在BASE上可以自由的调节LORA的强度, 但在turbo上生成是雪花 https://civarchive.com/models/2359268
纸飞机群交流 https://t.me/+gTE-veRY4qxkMDQx
Eason ZimageBASE workflow download
傻瓜提示词 for V3 download
(02-02-2016) updated：
I saw a user's comment today saying that after using my checkpoint, some features become weak or stop working properly (like "panty stockings").
This usually happens because when you merge a LoRA into a checkpoint, the merging process forces a fixed weight.
If the weight is set too low → the LoRA effect (e.g. pussy, nipples, detailed underwear etc.) becomes very subtle or almost invisible.
If the weight is set too high → it starts to override and damage the base model's original style/quality.
To solve this problem properly, I released this adjustable-weight version of the LoRA. You can now freely control the LoRA strength yourself — just lower it a bit if it's too strong, or increase it if you want the effect more obvious. Adjust it to whatever best matches your taste and prompt!
https://civarchive.com/models/2359268?modelVersionId=2653444
Hope this helps~ Feel free to experiment with different strengths (0.7–1 is usually a good range to play with) 😊
Merge checkpoint lets you handle some NSFW scenes more easily on their own — without needing extra LoRAs. You can get stuff like bondage outfit in the rain or wet micro bikini just from the merged model, skipping the hassle of loading several LoRAs at once.
今天看见一个网友留言说使用了我的Checkpoint，有些功能会失效，比如panty stockings。这是因为merge checkpoint选择融合LORA的时候会强制选择权重，权重低的时候，LORA本身数据会不明显，比如pussy，nipples。当权重强时候，会影响到原始基座模型的数据，所以为了解决这个问题，推出了这个可以调节权重的LORA，你可以搭配你的需要来适度增减LORA的强度，达到你的要求。https://civarchive.com/models/2359268?modelVersionId=2653444
Eason Z-Image-Turbo Realistic NSFW V2 (full bf16) (01-30-2026)
Model Update / V3 Preview
I tested the checkpoint yesterday and generated a bunch of example images. What I found: scenes with two-person action/poses (e.g., couple interactions, dynamic duo NSFW) tend to break down or collapse easily. This is likely because the overall dataset is quite large, but the specific training data for multi-character / double-person dynamic scenes wasn't sufficient — in short, the model simply didn't learn those well enough.
To address this, I'm planning to release V3 as an experimental attempt. No guarantees it will fully fix everything — I've done my best, but training outcomes can be unpredictable.
What changes in V3?
I'll completely remove all clothing-related tags/units and any general "AI-style" or clothed data.
The entire training focus will shift to double-person action / two-character interactions (especially NSFW couple poses, movements, and dynamics).
Recommendation for users:
If you're mainly into regular / solo NSFW, V2 is already the definitive, all-rounded version — it's a "best-of" compilation. You can generate tons of fun, high-quality images without needing any LoRA at this point. Stick with V2 for everyday use!
If you specifically love two-person / couple action scenes, hang tight for V3. Training will take about 20 days, and I'm hoping it delivers solid, improved NSFW results in that area.
Fingers crossed V3 turns out great — I'll share progress when it's closer. Thanks for the support, and feel free to drop any feedback or suggestions! 🚀
昨天测试了模型，生成了许多示例图，发现双人运动的时候图片容易崩，应该是数据量太大，而双人运动的图学习不够，简单说没学好。所以为了解决这个问题将会推出V3版本，但是这是尝试，也不一定解决，我也尽力了。V3会剔除所有的衣服和AI图相关的单元，把学习资源重点放在双人运动上，所以如果你想玩普通的NSFW那V2版本已经是集大成之作，不需要LORA已经可以生成很多好玩的图。但是喜欢双人运动的话就要等训练20天的V3模型了，希望V3版本能有不错的NSFW的效果。
纸飞机群交流
https://t.me/+gTE-veRY4qxkMDQx
Eason ZITurbo realistic NSFW fp8-pruned V2 (01-30-2026)
checkpoint type: z-image-turbo-fp8-e4m3fn
training checkpoint: z_image_turbo_bf16.safetensors
For detailed example images, please refer to the page for the Z-Image Turbo bf16 v2 version; the prompts are universal
千呼万唤屎出来，之前有人留言说要这个FP8版本，所以照顾一下小显存的用户推出了，提示词在full BF16 V2版本的页面里照搬，提示词是通用的。画质差点点，手指有崩坏的概率，私处就更难说了，没办法不是我的问题，是基座模型的问题，有条件还是用full BF16的吧
纸飞机群交流
https://t.me/+gTE-veRY4qxkMDQx
updated:
Eason Z-Image-Turbo Realistic NSFW V2 (full bf16) (01-29-2026)
I will update the example images later.
Eason Z-Image-Turbo Realistic NSFW V2 (Name may be adjusted later) (1/19/2026)
This model is currently in training (V2 version). It has already reached an insane 1,000,000 training steps using 9,432 carefully curated high-quality images, with an estimated total training time of around 15 days of non-stop computation! (10 more day from1/19/2026)
V2 fully preserves everything great from V1 while adding roughly 2x more new data for dramatically improved NSFW quality, diversity, and fidelity. Once released, the model page will include plenty of showcase generations demonstrating the new enhancements.
Version Comparison Table (Data is from ongoing training — final results, effects, and stats may change upon completion):
Important Note: This is a work-in-progress version. Final performance, trigger reliability, and any potential issues (e.g., overfitting or new data biases) may still be refined before release.
Stay tuned for the epic evolution in V2 — can't wait to drop it! 🔥 Once ready, download, test, and share your creations!
Eason Z-Image-Turbo Realistic NSFW V2 （名称可能后续调整）
这款模型目前正在训练中（V2 版本），训练步数已飙升至惊人的 1,000,000 步，使用 9432 张 高质量精心挑选图片，理论总历时约 15 天 不间断训练！(剩10 天 2026/1/19)
V2 在完整保留 V1 所有核心优势的基础上，新增了约 2 倍的全新数据，大幅提升 NSFW 表现力和多样性。模型发布后，将附上大量生成示例图，展示新增强的内容。
版本对比表格（当前数据为训练中状态，实际效果与最终数据可能会有变动）：
注意：模型仍在训练阶段，最终效果、触发词稳定性及潜在问题（如过拟合或新数据引入的偏差）都可能调整。敬请期待正式发布版本！
尽情期待 V2 的惊艳进化～ 训练完成后，欢迎下载测试并分享你的神图！
纸飞机群交流
https://t.me/+gTE-veRY4qxkMDQx
Eason Z-Image-Turbo Realistic NSFW V1
This merged checkpoint was forged from over 3000 carefully curated high-quality images, trained for 300,000 steps across 5 intense days and nights — built specifically to deliver superior, ultra-seductive NSFW body results!
Key Highlights:
Dramatically enhanced physique: larger, fuller breasts, thicker, more powerful thighs, perfect curvy hourglass figure with optimized proportions for maximum allure and realism!
Seamlessly integrates elements from my previous popular LoRAs: rope bondage , bondage suit, micro bikinis, and other fetish details trigger naturally and effortlessly.
Infused with realistic Tifa Lockhart features — trigger high-fidelity Tifa NSFW renders with a single prompt, now upgraded with the enhanced voluptuous body for breathtaking impact!
Beyond the showcase examples, this model holds even more hidden potential — waiting for your creativity to discover and unleash!
Dive in, explore her endless possibilities, and create something extraordinary! If you love it, drop a ❤️ + download, and share your masterpieces in the comments! 🔥
Someone asked where my workflow is — actually, every example image that's in PNG format has its own independent workflow embedded in the metadata.
Just drag and drop any of my showcase PNG images directly into your ComfyUI interface, and you'll instantly load the exact workflow I used for that image. Super simple! 🚀
Eason Z-image-Turbo realistic NSFW V1
Checkpoint type: ZIT
trigger：tifa lockhart, zitti, red eyes, long hair, breasts, solo, long hair, black hair, jewelry, sitting, earrings, cleavage, looking at viewer, dress, couch, lamp, white dress, indoors, realistic, long sleeves, blurry, collarbone, lips, medium breasts, blurry background, short dress, pillow, closed mouth
base checkpoint: z_image_turbo_bf16.safetensors
Total Training Steps: 300,000
Dataset Size: 3,000 images
Mixed Precision: bf16
Transformer Precision: float8 (FP8)
Built-in VAE: No
Built-in Text Encoder/CLIP: No
LoRA Loading Supported: Yes
Recommended Sampler: DDIM or Euler
Recommended Scheduler: simple or sgm_uniform
Denoising Strength: 1.0 (full generation)
中文名: 淫模_V1
这款融合模型（Checkpoint）由 3000+ 张精心挑选的高质量图片 历经 30万步训练、耗时 5天5夜 不眠不休打磨而成（真是苦了咱家的4090显卡😂），专为生成更极致、更诱人的 NSFW body 而生！
核心亮点：
显著强化 身材曲线：更大更饱满的乳房、更粗壮有力的腿部、完美 curvy 沙漏型身材，整体比例更性感、更写实！
继承并融合了我以往多款经典 LoRA 元素：绳索捆绑（shibari/rope bondage）、束缚制服、迷你比基尼（micro bikini） 等 fetish 细节自然触发，轻松出图！
特别注入 Tifa（蒂法）写实风格 元素，一句提示词就能直接生成高还原度的 Tifa NSFW 版本，结合优化后的丰满身材，视觉冲击拉满！
除了示例图展示的姿势与场景，模型还隐藏了更多惊喜潜力——等待你用创意去发掘与创造！
尽情探索她的无限可能吧～ 喜欢的话点个 ❤️ + 下载，期待你的神图分享！
纸飞机群交流
https://t.me/+gTE-veRY4qxkMDQx
