ComfyUI_LucidFlux

ComfyUI_LucidFlux

LucidFlux: Caption-Free Universal Image Restoration with a Large-Scale Diffusion Transformer，you can use it in ComfyUI

Update

• supported PID 4K super-resolution
• 新增Pid 4K超分,选择2kto4k 及对应模型，2k模式效果一般，用于512 to 2048

1.Installation

In the ./ComfyUI/custom_nodes directory, run the following:

git clone https://github.com/smthemex/ComfyUI_LucidFlux

2.requirements

pip install -r requirements.txt

3.checkpoints

• any flux dit / 任意flux模型， KJ的 或者官方封装的

• lucid checkpoints/ links /lucidflux.pth 和prompt_embeddings.pt 和gemma_prompt_embedding.pt

• siglip512 links / model.safetensors 只下单体模型

• DiffBIR links / general_swinir_v1.ckpt

• turbo lora links #optional 可选，8 步起

• lucid_connector.pthlinks

• flux ae links #

• ultravaelinks ,rename or not

• PIDlinks / PiD_res2kto4k_sr4x_official_flux_distill_4step.pth or PiD_res2k_sr4x_official_flux_distill_4step.pth

• null_caption_embs.ptlinks #PID use

├── ComfyUI/models/
|     ├── diffusion_models/any flux dit # 任意flux dit模型 ，就用kj的或者x flux的，名字要带dev 否则跑schnell
|     ├── diffusion_models
|        ├──PiD_res2kto4k_sr4x_official_flux_distill_4step.pth or PiD_res2k_sr4x_official_flux_distill_4step.pth # PID use
|     ├── vae/ae.safetensors #comfy 
|     ├── clip_vision/siglip2-so400m-patch16-512.safetensors  #rename from model.safetensors  最好重命名个，不然都是siglip 的model.safetensors
|     ├── LucidFlux/
|        ├──general_swinir_v1.ckpt
|        ├──lucidflux.pth
|        ├──prompt_embeddings.pt # 已适配，使用时不要连clip
|        ├── lucid_connector.pth
|        ├──null_caption_embs.pt # PID use
|        ├──gemma_prompt_embedding.pt # PID use

4 Example

• 4K
• 2k
• 1k

5 Citation

---

• FLUX
• X-flux
• PID

LucidFlux

@article{fei2025lucidflux,
  title={LucidFlux: Caption-Free Universal Image Restoration via a Large-Scale Diffusion Transformer},
  author={Fei, Song and Ye, Tian and Wang, Lujia and Zhu, Lei},
  journal={arXiv preprint arXiv:2509.22414},
  year={2025}
}