ComfyUI
ComfyUI copied to clipboard
The most powerful and modular diffusion model GUI, api and backend with a graph/nodes interface.
 ### Your question I've downloaded VisualStudio, CUDA Toolkit by following some videos also added commands but all guides for ComfyUI Portable, not desktop. Now I get this error. ###...
### Feature Idea gfx12 (9060, 9070) brings support for fp8 on AMD consumer cards. This functionality is currently soldered off, I think by comfy/model_management.py, if not other parts of the...
BE change for https://github.com/Comfy-Org/ComfyUI_frontend/pull/3976 Add one fetch node. Why do we need this node? In certain situations, such as in cloud environments (RunningHub, ComfyDeploy, etc.), users cannot directly access the...
### Expected Behavior "Load video“node in Comfy Core cannot be used normally,After selecting, there is no response and the node cannot be loaded.  ComfyUI version is 0.3.35,frontend is 1.20...
  ### Getting 'IndexError: list index out of range' in ComfyUI - Help with LoRA Caption Save Node  Hi all, I’m using ComfyUI on a Mac Studio (M1...
### Your question # ComfyUI Error Report ## Error Details - **Node ID:** 17 - **Node Type:** WanVideoImageClipEncode - **Exception Type:** torch.cuda.OutOfMemoryError - **Exception Message:** Allocation on device ## Stack...
### Expected Behavior In the latest version, the "Convert widget to input" function for the CLIP node is missing. I've tested multiple versions, and it worked fine in versions prior...
### Feature Idea Arch Linux requieres to set: `export CMAKE_POLICY_VERSION_MINIMUM=3.5 ` ` export CXXFLAGS="-include cstdint" ` Due to Sentencepiece, the installation process on Arch Linux would encounter compilation errors without...
### Your question **Environment:** - GPU: RTX 5090 - CUDA: 12.8 - PyTorch: 2.7.0 **Question:** Has anyone successfully implemented video dubbing workflows in ComfyUI? I'm looking for guidance on: 1....
### Expected Behavior what's wrong?  but lora can still be used in V0.3.29 ### Actual Behavior what's wrong?  but lora can still be used in V0.3.29 ### Steps...