Premium character LoRAs
Original synthetic characters, bundles and AI tools are in my shop:
Join the community on Discord: https://discord.gg/nB79wRNtMT
Tight Dresses
A concept LoRA for Krea 2 focused on tight, body-hugging dresses, NSFW: bodycon, ribbed knit, satin and latex dresses that cling to every curve. It gives an amateur photo look.
Usage
Start with the trigger word TightDressRK, then describe the dress (bodycon, sweater dress, satin, latex, ribbed knit; mini or midi; neckline, straps, color), how it clings and what it outlines (breasts, nipples through the fabric, hips, buttocks), her body, hair and shoes, her pose (standing, turned to show her back, mirror selfie), the place (night out, street, bedroom), and end with the light. Describing how tightly the fabric clings brings out the effect the most.
TightDressRK, Photograph of a woman in <a tight red ribbed knit mini dress clinging to her hips>, <black heels>, <turned to show her back, mirror selfie> in a <hallway>. The lighting is <warm evening light>.
Recommended weight: 1. Made for Krea 2 Turbo (8 steps, CFG 1).
Prompt Examples
"TightDressRK, Photograph of an adult woman with a light skin tone and long, straight brown hair with bangs. She has a curvy, hourglass figure with prominent buttocks and visible tattoos on her left thigh and right arm."
"TightDressRK, Photograph of a woman with light olive skin and long, straight brown hair. She wears black-rimmed glasses and a tight, blue, spaghetti-strap, bodycon dress that clings to her curvy, hourglass figure, outlining her large breasts, narrow waist, and wide hips."
"TightDressRK, Photograph of an adult woman with fair skin and long, straight blonde hair. She has a curvy, hourglass figure with large breasts and a prominent, rounded buttocks."
Training
- Base model: Krea 2
- Type: Concept LoRA
- Steps: 4000
- Rank: 32
- Optimizer: adamw, constant learning rate
- Engine: ai-toolkit
Generation settings
Trained on Krea 2, and works on Krea 2 Turbo for faster generation.
- Raw: ~30 steps, guidance 3.5
- Turbo: ~8 steps, guidance 1.0
The gallery images were generated with krea2_turbo_mxfp8.safetensors, at 8 steps, guidance 1, LoRA weight 1.



















