🐰 BUNNY’s Heavy NSFW Wildcards| Solo Only | LoRA Friendly

FourBunnyFourBunny
5
114
poseswildcardsnsfwfluxqwenzimagekrea2
part1
model-image-0
model-image-1
model-image-2
model-image-3
model-image-4
model-image-5
model-image-6
model-image-7
model-image-8
model-image-9
model-image-10
model-image-11
model-image-12
model-image-13
model-image-14
model-image-15
model-image-16
model-image-17
model-image-18
model-image-19

Details

Download Files (1)

Model description

🐰 BUNNY’s NSFW Wildcards

Heavy/Hardcore NSFW Reworked Version | Solo | Character LoRA Friendly

🔞 About This Version / 关于此版本

This is the refined Heavy NSFW version of my wildcard series, specifically optimized for character LoRA generation. The goal is a controlled NSFW fine-tune on the source text, not a full scene rewrite. It preserves the original scene, composition, camera angle, lighting, and pose foundation as much as possible, while pushing the content into a stronger solo NSFW direction (e.g., Nude, Half-nude, Open/Torn clothing, Displaced outfit state). Heavy NSFW, BDSM, and fetish elements may be fused when suitable.

这是我 Wildcard 系列中的精修重度 NSFW 版本，专门针对人物 LoRA 出图进行优化。其核心目标是在原文基础上做可控的 NSFW 微调，而非整段推倒重写。它会最大程度保留原始场景、构图、镜头、光影、氛围和基础姿势，同时将内容推向更强的单人 NSFW 方向（如：全裸、半裸、敞开/撕破服装、拉开/错位穿着等）。在合适的情况下融合 Heavy NSFW、BDSM 与 Fetish 风格元素。

📦 Release Info & Credits / 规模与致谢

  • 📌 Original Post / 原帖地址: SFW15000 & Qwen Wildcards — Huge thanks to the original author for creating this detailed prompt collection!

  • 感谢原作者制作了如此庞大且细节丰富的 Prompt 合集！🙏

  • 📊 Current Size (Part 1): Contains 1,663 different prompts (~1.45M characters / 360K tokens). The remaining two parts are of similar size and will be released after testing.

  • 📊 当前发布规模 (第一部分): 包含 1,663条 不同提示词（约 145万字符 / 36万 tokens）。剩余两部分体量基本相同，测试完成后陆续发布。

🎨 Core Rules & Logic / 核心规则与逻辑

  • 🐰 Character LoRA Friendly / 人物 LoRA 友好设计

  • To avoid fighting against your character LoRA, this version intentionally removes all facial features, face shapes, eye/nose/lip details, makeup, skin details, hairstyles, hair colors, hair lengths, and bangs/ponytail descriptions. The wildcard handles the scene, outfit state, camera, and NSFW action, while your LoRA safely controls the face, hair, and character identity.

  • 为了避免与人物 LoRA 冲突，本版本刻意去除了所有五官、脸型、眼鼻唇、妆容、发型、发色、发长及刘海等细节描写。Wildcard 仅负责场景、姿势、服装状态与 NSFW 动作，让 LoRA 完美掌控角色面部与身份特征。

🔥 Content Rules / 内容规则

  • This version is solo only.

  • 本版本为 纯单人 SOLO。

  • The NSFW actions mainly focus on solo scenarios, including self-play, fingers, toys, vibrators, spreading, solo climax, and fluid-related details.

  • NSFW 动作主要围绕单人场景，包括自慰、手指、玩具、震动棒、自己扒开、单人高潮和体液细节等。

  • Heavy NSFW, BDSM, and fetish-style elements may be lightly fused into the original scene when appropriate.

  • 在适合的情况下，会在原始场景中轻度融合 Heavy NSFW、BDSM 与 Fetish 风格元素。

  • Each line aims to include visible adult exposure and at least one fluid or mess detail, while still preserving the original scene logic as much as possible.

  • 每条 Prompt 尽量包含明确的成人暴露内容，并加入至少一种液体或凌乱细节，同时尽可能保留原始场景逻辑。

  • 👗 Clothing State / 服装穿着状态

  • the original outfit is usually preserved and shifted into compromised states: Nude, half-nude, torn clothing, open clothing, or displaced outfit states. This preserves the original character setup instead of turning every prompt into the same generic NSFW outfit.

  • 通常保留原始服装并改变穿着状态（全裸、半裸、撕破、敞开或错位状态），以此保留原角色的设定与场景风格，避免千篇一律。

🧍 Why Solo Only? / 为什么仅限单人 SOLO？

This version is adult solo only (no partners/multi-character interaction). It was processed via my custom custom API translation/rewriting pipeline using DeepSeek Flash. During testing, multi-person interactions often led to messy limb errors or broken contact logic due to prompt complexity. Rewriting it into a solo-only format vastly improves stability and practical usability.

本版本为成人单人 SOLO 专用（不含搭档或多人互动）。整个合集基于 DeepSeek Flash 通过我自制的高并发 API 流程完成“中文改写+英文回译”的双阶段处理。测试中发现多人互动提示词极易导致肢体错误或互动逻辑混乱，因此全面改为纯单人模式，以确保最高稳定性和实用性。

👤 Character LoRA Friendly / 人物 LoRA 友好设计

To avoid fighting against your character LoRA's identity, this version intentionally removes all:

❌ Facial-features, face-shapes, eye/nose/lip/skin details, makeup, and expressions.

❌ Hairstyles, hair-color, hair-length, bangs, ponytails, or braids. The wildcard handles the environment, clothing state, camera, and NSFW action—while your LoRA carries the character identity.

为了避免和人物 LoRA 产生严重的权重冲突，本版本刻意去除了所有：

❌ 五官/脸型描写、眼睛/鼻子/嘴唇/皮肤细节、妆容及绝大多数表情描写。

❌ 发型、发色、发长、刘海、马尾、辫子等一切头发细节描写。

Wildcard 仅全权负责场景、姿势、服装状态、镜头、光影、氛围和 NSFW 动作，让 LoRA 完美控制角色本尊的脸与特征。

🧪 Practical Guide / 使用与兼容性指南

  • ⚠️ Model Compatibility / 模型兼容性: Best suited for modern large models that support detailed scene descriptions and long natural-language prompting.

  • 更适合支持详细场景、镜头语言和长自然语言提示词的现代大型图像生成模型.

⚙️ How It Was Made / 制作方式

This version was created using my own custom API translation / rewriting pipeline.

这个版本使用我自制的 API 翻译 / 改写流程制作。

The workflow uses:

⚙️ High-concurrency API processing
📄 One source segment sent independently each time
🎲 A recipe pool for randomness
🧩 Different pose recipes
👗 Different clothing-state recipes
🔞 Different NSFW category recipes
🇨🇳 Chinese-side rewriting first
🇬🇧 Then translated back into English

流程包括：

⚙️ 高并发 API 处理
📄 每段文本单独独立发送
🎲 使用配方池保证随机性
🧩 不同姿势配方
👗 不同服装状态配方
🔞 不同 NSFW 类别配方
🇨🇳 先在中文侧完成修改
🇬🇧 再翻译返回英文

This helps keep each prompt independent, reduces repeated patterns, and gives the wildcard collection more random variation.

这样可以让每条 Prompt 保持独立，减少重复模式，并让整个 Wildcard 合集拥有更多随机变化。

💬 Feedback & Disclaimer / 反馈与说明

In my testing, around 5% of images may still show anatomy/object placement errors, but it is highly stable for batch runs. If you have unique ideas or prompt logic to improve the next parts, feel free to leave a comment or DM me! 🐰🔥

根据实际生图测试，约有 5% 的图可能存在肢体或道具位置异常，但对于批量跑图已足够稳定。如果你有更好的 Prompt 改进建议，欢迎留言或私信，有价值的反思会直接优化到后续章节中！玩得开心！🐰🔥

Images made by this model