2.5Dfantasycharacter - Wan T2V 14B

This LoRA is designed for generating WAN-based European fantasy–style game character videos, while integrating Eastern aesthetics to consistently produce anime-style characters with a Chinese wuxia spirit (such as swordsmen, female warriors, or hybrid knight × wuxia styles).

During training, a high-quality European fantasy character model was used as the base. The dataset consists primarily of stylistically consistent character illustrations, including half-body and full-body artwork. Special emphasis was placed on preserving costume structure, armor patterns, material rendering, lighting, and character proportions. Style-related keywords such as “Chinese style,” “wuxia,” and “anime” were included in the captions to guide effective style fusion.

Hey everyone, I've discovered a gem of an AI image and video platform: RunningHub! ComfyUI developers worldwide release hundreds of incredibly fun and useful AI applications here every day, covering both entertainment and work. Open the link: https://www.runninghub.ai/?inviteCode=rh-v1396 to register and receive 1000 RH coins, which can be used to generate tons of free images and videos!