Stable Diffusion · LoRA
ObjectTypo Redmond WAN2 T2V 14B LoRA
A LoRA for Wan2.2-T2V-14B that generates ObjectTypography — words and letters formed by environmental objects, trained on cinematic and artistic composition…
About this model
This LoRA adapts the Wan-AI/Wan2.2-T2V-A14B text-to-video model to produce ObjectTypography, a creative technique where words and letters are spelled out by natural and man-made elements arranged in the environment. The model forms words through crops at different growth stages, light beams, chalk trails, neon signs, rusting metal, or lightning patterns, turning environmental composition into readable letterforms. Two weights are provided: a high-noise and a low-noise variant, giving flexibility for different generation settings.
To use the model, start any prompt with the trigger phrase "ObjectTypography" followed by a description of how a word is physically formed in a scene. The model handles the compositional reasoning to arrange objects into legible letterforms while preserving cinematic style, lighting, and atmosphere.
Place the safetensors file in your ComfyUI `models/loras/` directory, then add a Load LoRA node after the base Wan model loader in your workflow. A starting strength of 0.8 to 1.0 is recommended for strong object-typography effect. The base model is Wan-AI/Wan2.2-T2V-A14B and the license is CC0 1.0.
This project was supported by reDMOND Ai and is part of the broader ObjectTypography series by ArtificialGuyBR.
Learn more
Read the in-depth guide about ObjectTypo Redmond WAN2 T2V 14B LoRA
Project signals
- 18 Hugging Face downloads
- 1 Hugging Face likes
Topics
diffusers · lora · wan · text-to-video · object-typography · creative · artistic · text-to-image · template:diffusion-lora · base_model:wan-ai/wan2.2-t2v-a14b