ComfyUI-MultiGPU

CheckpointLoaderAdvancedMultiGPU

The CheckpointLoaderAdvancedMultiGPU node is used to load checkpoint models (complete diffusion models containing UNet, CLIP, and VAE components) with granular device control, allowing individual placement of each model component on different GPUs or devices. This node automatically detects models located in the ComfyUI/models/checkpoints folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | ckptname | STRING | The name of the checkpoint model to load. | | unetdevice | STRING | Target device for the UNet diffusion model component (e.g., 'cuda:0', 'cuda:1', 'cpu'). | | clipdevice | STRING | Target device for the CLIP text encoder component (e.g., 'cuda:0', 'cuda:1', 'cpu'). | | vaedevice | STRING | Target device for the VAE decoder/encoder component (e.g., 'cuda:0', 'cuda:1', 'cpu'). | | Output Name | Data Type | Description | | --- | --- | --- | | MODEL | MODEL | The loaded UNet diffusion model. | | CLIP | CLIP | The loaded CLIP text encoder model. | | VAE | VAE | The loaded VAE decoder/encoder model. |

Multigpu · 3 outputs

CheckpointLoaderSimpleDisTorch2MultiGPU

The CheckpointLoaderSimpleDisTorch2MultiGPU node is used to load checkpoint models (complete diffusion models containing UNet, CLIP, and VAE components) with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/checkpoints folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | ckptname | STRING | The name of the checkpoint model to load. | | computedevice | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: true). | | Output Name | Data Type | Description | | --- | --- | --- | | MODEL | MODEL | The loaded UNet diffusion model with DisTorch…

CheckpointLoaderSimpleMultiGPU

The CheckpointLoaderSimpleMultiGPU node is used to load checkpoint models (complete diffusion models containing UNet, CLIP, and VAE components) with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node automatically detects models located in the ComfyUI/models/checkpoints folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | ckptname | STRING | The name of the checkpoint model to load. | | device | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | MODEL | MODEL | The loaded UNet diffusion model. | | CLIP | CLIP | The loaded CLIP text encoder model. | | VAE | VAE | The loaded VAE decoder/encoder model. |

clip_name1

Compact description unavailable for this node.

clip_name2

Compact description unavailable for this node.

clip_name3

Compact description unavailable for this node.

clip_name4

Compact description unavailable for this node.

CLIPLoaderDisTorch2MultiGPU

The CLIPLoaderDisTorch2MultiGPU node is used to load standard CLIP text encoder models with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger text encoding models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/clip folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipname | STRING | The name of the CLIP model to load. | | type | STRING | The type of CLIP model (e.g., 'stablediffusion', 'stablediffusionxl'). | | device | STRING | Target device for text encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: false for CLIP loaders). | | Output Name | Data Type | Description | | --- | --- | --- | | CLIP | CLIP…

CLIPLoaderMultiGPU

The CLIPLoaderMultiGPU node is used to load CLIP text encoder models with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node automatically detects models located in the ComfyUI/models/clip folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipname | STRING | The name of the CLIP model to load. | | type | STRING | The type of CLIP model (e.g., 'stablediffusion', 'stablediffusionxl'). | | device | STRING | Target device for text encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | CLIP | CLIP | The loaded CLIP text encoder model. |

CLIPVisionLoaderDisTorch2MultiGPU

The CLIPVisionLoaderDisTorch2MultiGPU node is used to load CLIP Vision models with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger vision encoder models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/clipvision folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipvision | STRING | The name of the CLIP Vision model to load. | | device | STRING | Target device for vision encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: true). | | Output Name | Data Type | Description | | --- | --- | --- | | CLIPVISION | CLIPVISION | The loaded CLIP Vision model with DisTorch2 distributed allocation applied. | DisTorc…

CLIPVisionLoaderMultiGPU

The CLIPVisionLoaderMultiGPU node is used to load CLIP Vision models with device selection capability, enabling users to specify which GPU or device should be used for vision encoder execution. This node automatically detects models located in the ComfyUI/models/clipvision folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipvision | STRING | The name of the CLIP Vision model to load. | | device | STRING | Target device for vision encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | CLIPVISION | CLIPVISION | The loaded CLIP Vision model. |

ControlNetLoaderDisTorch2MultiGPU

The ControlNetLoaderDisTorch2MultiGPU node is used to load ControlNet models with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger conditional generation models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/controlnet folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | controlnetname | STRING | The name of the ControlNet model to load. | | computedevice | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: true). | | Output Name | Data Type | Description | | --- | --- | --- | | CONTROLNET | CONTROLNET | The loaded ControlNet model with DisTorch2 distributed allocation applied. | DisTor…

ControlNetLoaderMultiGPU

The ControlNetLoaderMultiGPU node is used to load ControlNet models with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node automatically detects models located in the ComfyUI/models/controlnet folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | controlnetname | STRING | The name of the ControlNet model to load. | | device | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | CONTROLNET | CONTROLNET | The loaded ControlNet model. |

default

Compact description unavailable for this node.

DeviceSelectorMultiGPU

Compact description unavailable for this node.

Multigpu · 1 output

DiffControlNetLoaderDisTorch2MultiGPU

The DiffControlNetLoaderDisTorch2MultiGPU node is used to load Diffusers ControlNet models (HuggingFace Hub repositories) with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger conditional generation models across multiple GPUs. This node loads ControlNet models directly from HuggingFace model repositories by specifying the repository ID (e.g., "diffusers/controlnet-canny-sdxl-1.0"). | Parameter | Data Type | Description | | --- | --- | --- | | modelpath | STRING | The HuggingFace repository ID or local path of the diffusers ControlNet model to load. | | computedevice | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: true). | | Output Name | Data Type | Description | | --- | --- | --- | | CONTROLNET | CONTROLNET | The loaded diffusers ControlNet model with DisTorch2 distributed allocation applied. | DisTorch2 is an advanced memory management system that enable…

DiffControlNetLoaderMultiGPU

The DiffControlNetLoaderMultiGPU node is used to load Diffusers ControlNet models (HuggingFace Hub repositories) with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node loads ControlNet models directly from HuggingFace model repositories by specifying the repository ID (e.g., "diffusers/controlnet-canny-sdxl-1.0"). | Parameter | Data Type | Description | | --- | --- | --- | | modelpath | STRING | The HuggingFace repository ID or local path of the diffusers ControlNet model to load. | | device | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | CONTROLNET | CONTROLNET | The loaded diffusers ControlNet model. |

DiffusersLoaderDisTorch2MultiGPU

The DiffusersLoaderDisTorch2MultiGPU node is used to load Diffusers models (HuggingFace Hub repositories) with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger diffusion models across multiple GPUs. This node loads models directly from HuggingFace model repositories by specifying the repository ID (e.g., "stabilityai/stable-diffusion-xl-base-1.0"). | Parameter | Data Type | Description | | --- | --- | --- | | modelpath | STRING | The HuggingFace repository ID or local path of the diffusers model to load (e.g., 'stabilityai/stable-diffusion-xl-base-1.0'). | | computedevice | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: true). | | Output Name | Data Type | Description | | --- | --- | --- | | MODEL | MODEL | The loaded diffusers model with DisTorch2 distributed allocation applied. | DisTorch2 is an advanced memory management system that enables loading and ru…

DiffusersLoaderMultiGPU

The DiffusersLoaderMultiGPU node is used to load Diffusers models (HuggingFace Hub repositories) with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node loads models directly from HuggingFace model repositories by specifying the repository ID (e.g., "stabilityai/stable-diffusion-xl-base-1.0"). | Parameter | Data Type | Description | | --- | --- | --- | | modelpath | STRING | The HuggingFace repository ID or local path of the diffusers model to load (e.g., 'stabilityai/stable-diffusion-xl-base-1.0'). | | device | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | MODEL | MODEL | The loaded diffusers model. |

DualCLIPLoaderDisTorch2MultiGPU

The DualCLIPLoaderDisTorch2MultiGPU node is used to load dual standard CLIP text encoder models with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger text encoding models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/clip folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipname1 | STRING | The name of the first CLIP model to load. | | clipname2 | STRING | The name of the second CLIP model to load. | | type | STRING | The type of CLIP model configuration for dual loading. | | device | STRING | Target device for text encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: false for CLIP loaders). | | Outpu…

DualCLIPLoaderMultiGPU

The DualCLIPLoaderMultiGPU node is used to load dual CLIP text encoder models with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node automatically detects models located in the ComfyUI/models/clip folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipname1 | STRING | The name of the first CLIP model to load. | | clipname2 | STRING | The name of the second CLIP model to load. | | type | STRING | The type of CLIP model configuration for dual loading. | | device | STRING | Target device for text encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | CLIP | CLIP | The loaded dual CLIP text encoder models. |

multiline

Compact description unavailable for this node.

QuadrupleCLIPLoaderDisTorch2MultiGPU

The QuadrupleCLIPLoaderDisTorch2MultiGPU node is used to load quadruple standard CLIP text encoder models with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger text encoding models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/clip folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipname1 | STRING | The name of the first CLIP model to load. | | clipname2 | STRING | The name of the second CLIP model to load. | | clipname3 | STRING | The name of the third CLIP model to load. | | clipname4 | STRING | The name of the fourth CLIP model to load. | | device | STRING | Target device for text encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for test…

QuadrupleCLIPLoaderMultiGPU

The QuadrupleCLIPLoaderMultiGPU node is used to load quadruple CLIP text encoder models with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node automatically detects models located in the ComfyUI/models/clip folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipname1 | STRING | The name of the first CLIP model to load. | | clipname2 | STRING | The name of the second CLIP model to load. | | clipname3 | STRING | The name of the third CLIP model to load. | | clipname4 | STRING | The name of the fourth CLIP model to load. | | device | STRING | Target device for text encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | CLIP | CLIP | The loaded quadruple CLIP text encoder models. |

TripleCLIPLoaderDisTorch2MultiGPU

The TripleCLIPLoaderDisTorch2MultiGPU node is used to load triple standard CLIP text encoder models with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger text encoding models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/clip folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipname1 | STRING | The name of the first CLIP model to load. | | clipname2 | STRING | The name of the second CLIP model to load. | | clipname3 | STRING | The name of the third CLIP model to load. | | device | STRING | Target device for text encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: false for CLIP loaders). | | Output Na…

TripleCLIPLoaderMultiGPU

The TripleCLIPLoaderMultiGPU node is used to load triple CLIP text encoder models with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node automatically detects models located in the ComfyUI/models/clip folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | clipname1 | STRING | The name of the first CLIP model to load. | | clipname2 | STRING | The name of the second CLIP model to load. | | clipname3 | STRING | The name of the third CLIP model to load. | | device | STRING | Target device for text encoder compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | CLIP | CLIP | The loaded triple CLIP text encoder models configured for SD3. |

type

Compact description unavailable for this node.

UNETLoaderDisTorch2MultiGPU

The UNETLoaderDisTorch2MultiGPU node is used to load UNet diffusion models with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/unet folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | unetname | STRING | The name of the UNet model to load. | | computedevice | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: true). | | Output Name | Data Type | Description | | --- | --- | --- | | MODEL | MODEL | The loaded UNet model with DisTorch2 distributed allocation applied. | DisTorch2 is an advanced memory management system that enables lo…

UNetLoaderLP

UNetLoaderLP is a low-precision variant of the standard UNet loader that disables high-precision LoRA tensors for CPU-stored models, freeing additional host memory while remaining compatible with MultiGPU device routing. | Parameter | Data Type | Description | | --- | --- | --- | | unetname | STRING | UNet checkpoint filename from ComfyUI/models/unet. | | Parameter | Data Type | Description | | --- | --- | --- | | device | STRING | Device that should serve the UNet after loading. | | Output Name | Data Type | Description | | --- | --- | --- | | model | MODEL | Loaded UNet model with low-precision LoRA flag set. |

Loaders · 0 inputs · 1 parameter · 1 output

UNETLoaderMultiGPU

The UNETLoaderMultiGPU node is used to load diffusion model UNet components with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node automatically detects models located in the ComfyUI/models/unet folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | unetname | STRING | The name of the UNet model to load. | | device | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | MODEL | MODEL | The loaded UNet diffusion model. |

VAELoaderDisTorch2MultiGPU

The VAELoaderDisTorch2MultiGPU node is used to load VAE (Variational Autoencoder) models with DisTorch2 distributed tensor allocation, enabling advanced multi-device VRAM management to handle larger models across multiple GPUs. This node automatically detects models located in the ComfyUI/models/vae folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | vaename | STRING | The name of the VAE model to load. | | computedevice | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | virtualvramgb | FLOAT | Amount of virtual VRAM in gigabytes to allocate for distributed tensor management (default: 4.0, range: 0.0-128.0). | | donordevice | STRING | Device to donate VRAM from when allocating virtual memory (default: 'cpu'). | | expertmodeallocations | STRING | Advanced allocation string for expert users to manually specify device/ratio distributions (e.g., 'cuda:0,50%;cpu,'). | | ejectmodels | BOOLEAN | Whether to unload ALL models from the target device before loading this model, enabling deterministic model eviction for testing and memory management (default: true). | | Output Name | Data Type | Description | | --- | --- | --- | | VAE | VAE | The loaded VAE decoder/encoder with DisTorch2 distributed allocation applied. | DisTorch2 is an advanced memory management system…

VAELoaderMultiGPU

The VAELoaderMultiGPU node is used to load VAE (Variational Autoencoder) models with device selection capability, enabling users to specify which GPU or device should be used for model execution. This node automatically detects models located in the ComfyUI/models/vae folder, and it will also read models from additional paths configured in the extramodelpaths.yaml file. Sometimes, you may need to refresh the ComfyUI interface to allow it to read the model files from the corresponding folder. | Parameter | Data Type | Description | | --- | --- | --- | | vaename | STRING | The name of the VAE model to load. | | device | STRING | Target device for compute operations (e.g., 'cuda:0', 'cuda:1', 'cpu'). Selected from available devices on your system. | | Output Name | Data Type | Description | | --- | --- | --- | | VAE | VAE | The loaded VAE decoder/encoder model. |