Popular repositories Loading

1. ComfyUI-UniversalIO ComfyUI-UniversalIO Public

   Save/Load/Pass universal nodes for CONDITIONING & LATENT. Two-stage workflows for big models: batch encode → cloud sampling. Fixes path & XPU bugs; device auto/cpu/gpu; mode-based UI.

   Python

2. ComfyUI-XPU-CacheClean ComfyUI-XPU-CacheClean Public

   XPU/CUDA VRAM cache cleaner (soft/hard) for Intel Arc. Prevents level_zero OOM (UR_RESULT_ERROR_OUT_OF_RESOURCES).

   Python

3. HQH-XPU HQH-XPU Public

   HQH-XPU: Intel Arc A770 ComfyUI optimization bundle — TINT4 (oneDNN hybrid), OmniXPU, XPU-CacheClean, ClipProj (XPU) & UniversalIO; one install auto-deploys all nodes.

   Python

4. ComfyUI-TINT4 ComfyUI-TINT4 Public

   Forked from JWLHS/ComfyUI-TINT4

   TINT4 — torchao INT4 量化推理插件 for ComfyUI 基于 PyTorch 官方 torchao，将 DiT 扩散模型量化为 INT4 精度。

   Python

5. ComfyUI-OmniXPU ComfyUI-OmniXPU Public

   Forked from Jasonzzt/ComfyUI-OmniXPU

   Python

6. ComfyUI-ClipProj ComfyUI-ClipProj Public

   Forked from nicolab28/ComfyUI-ClipProj

   Swap a large text encoder for a small one with a learned linear projection. MiniMax H3 conditioning from 15.7 GB down to 5.2 GB.

   Python