Sitting Upskirt Edit LoRA for Qwen Image 2.1

A specialized edit model that transforms your AI photo into a natural "sitting upskirt" view. Simply upload an image of a woman sitting, describe the desired skirt/dress and panties (color, style, lace/regular, etc.), and the model will seamlessly generate the result without masks or additional LoRAs.

✨ Key Features

• Smart Edit: Upload a seated reference image and prompt for exact clothing details. The model understands context and transforms the view accurately.

• T2I Support: Generate sitting upskirt scenes directly from text prompts without any reference image.

• Ref2I Capability: Specify or reference exact skirt, dress, or pantie models for precise styling control.

• Optimized Resolution: Performs best at 2048×2048 but works reliably at lower resolutions as well.

• Zero Dependencies: Trained with a purpose-built dataset specifically for the default ComfyUI workflow of Qwen Image 2.1. No extra nodes or complex setups required.

• Turbo Compatible: Fully supports both standard and turbo versions of Qwen Image 2.1.

⚙️ Recommended Settings

• Use the default ComfyUI template for Qwen Image 2.1

• CFG 1 works fine, if you want more detail you can push up to CGF 3.

• Drop the LoRA into your workflow, load your reference image (or input a prompt), apply your desired settings, and generate.

☕ Support My Work

If you find this model useful, consider buying me a coffee to help fund future projects:
🔗 https://buymeacoffee.com/itwasfake

Thank you for using the model!