Pick your character LoRA, swap in your trigger word, and get a full character sheet in a single image: 4 turnaround views on the top row and 4 expressions on the bottom.

Includes

• A clean ComfyUI workflow with a read-me note inside

• Full quality mode (30 steps, about 6 min on a 16 GB GPU) or fast mode (Lightning 4 steps, about 20 seconds)

• Works with any Qwen-Image-2512 character LoRA. The examples use our free Detective Crumpet LoRA.

How to use: pick your LoRA in the green node, replace crump3t in the prompt with your trigger word, describe the outfit, and click Queue.

Models needed (all Apache 2.0): qwen_image_2512_fp8_e4m3fn, qwen_2.5_vl_7b_fp8_scaled, qwen_image_vae, and optionally Qwen-Image-2512-Lightning-4steps.