✨ Body– Figure Enhancer for FLUX 3 Klein ✨
Adds a fuller, curvier and more defined figure to women generated with FLUX 3 Klein, while keeping the face, pose, outfit and background intact. Works for text-to-image and image editing (reference) workflows.
🔥 What it does
- Fuller bust, narrower waist, wider hips and thicker thighs
- Natural skin texture, no plastic or waxy look
- Keeps identity, pose and scenery in edit workflows
- Works with swimwear, fitted clothing, standing or seated poses
☕ Support my work
If you like this LoRA and want more releases, buy me a coffee:
🔑 Trigger word
b!gb0dy → put it at the start of your prompt
⚙️ Recommended settings
- Strength: 0.8 – 1.0
- Sampler: Euler | Scheduler: Simple
- Steps: 25 | CFG: 3.5
- Resolution: ~2 MP (1440×1440 tested)
- Text encoder: Qwen3-VL 4B | VAE: flux2-vae
📝 Example prompt
b0d!, Full-color outdoor photograph, shot on natural daylight with soft, diffused ambient light casting even illumination across the scene; camera positioned at eye level, medium-wide framing capturing full-body posture with shallow depth of field blurring the background into textured wooden planks. In the foreground, a young Caucasian woman, mid-20s, stands in a deep lunge stretch pose — right knee bent deeply forward, left leg extended back with foot flat on ground — facing slightly rightward with focused, determined expression. She has fair skin, athletic build, blonde hair tied high in a sleek ponytail, and wears a black seamless sports bra and vibrant crimson red high-waisted leggings that cling to her toned physique, paired with dark gray/black trail running shoes. Her left hand rests lightly on her left thigh while her right arm hangs relaxed by her side. Behind her, the midground consists of weathered vertical wooden beams forming a rustic barn door backdrop, occupying roughly one-sixth of the frame as a muted brown-gray silhouette. On the left, her left foot is planted firmly on gritty dirt ground, while the right heel lifts slightly off the earth. No text or branding is present.
💡 Tips
- Use "full-body" or "standing" in the prompt for the strongest effect
- Lower to 0.6 – 0.7 for a subtler change
- In edit workflows add: "keep the same face, pose and background"
⚠️ Limitations
- Strong settings may slightly change how clothing fits
- Not tested on male subjects or anime/stylized models
- Intended for adult subjects only
🖼️ Comparisons
Each pair uses the same seed and prompt: first image without LoRA, second with LoRA at 1.0.