Stable Diffusion · LoRA

ThunderHero LoRA for Wan 2.2 Image-to-Video

LoRA for Wan 2.2 I2V that adds cinematic lightning, electricity, and superhero power effects to any input image via a two-file low/high-noise setup.

About this model

ThunderHero is a LoRA adapter for the Wan 2.2 Image-to-Video 14B model that specializes in electrifying superhero effects. When prompted with the trigger phrase "Thunder Hero," it transforms static input images into dynamic video sequences filled with lightning bolts, electrical arcs, and dramatic power manifestations. The effect works on ordinary scenes — portraits, cityscapes, objects — and reinterprets them as cinematic superhero moments.

The LoRA ships as two safetensors files: a low-noise variant and a high-noise variant, both loaded simultaneously in ComfyUI alongside the base Wan 2.2 I2V A14B model. The recommended strength_model range is 0.8 to 1.0. This two-file setup matches Wan 2.2's native high-noise/low-noise denoising pipeline, so both files need to be stacked via a Load LoRA node between the base model and KSampler.

The model is published under the CC0 1.0 public-domain dedication, making it free to use, modify, and redistribute without restriction. It was trained with GPU sponsorship from Redmond Ai. At 146 MB, it's a compact drop-in that pairs exclusively with Wan-AI/Wan2.2-I2V-A14B — it is not a standalone video model.