ComfyUI-AReA

ComfyUI-AReA

AReA is a training-free attention mechanism for generating images beyond the training resolution of diffusion transformers

Tip

• Support flux/qwen-image/ideogram now

1.Installation

git clone https://github.com/smthemex/ComfyUI-AReA

2.Checkpoints

normal comfyUI dit vae and te comfy-org
turbo lora flux turbo

3.Example

4.Citation

@misc{rajabi2026area,
      title={AReA: Adaptive Receptive-Field Attention for Efficient High-Resolution Diffusion Transformers},
      author={Javad Rajabi and Koorosh Roohi and David B. Lindell and Babak Taati},
      year={2026},
      eprint={2610.17227},
      archivePrefix={arXiv},
      primaryClass={cs.CV},
      url={https://arxiv.org/abs/2610.17227},
}