ComfyUI Node
đź’€Prepack Model SingleCLIP
Load a diffusion UNet and a single CLIP in one step (types follow official Load CLIP node).
đź’€Prepack Model SingleCLIP
- model
- clip
- vae
â—„unet_nameâ–ľâ–ş
â—„weight_dtypeâ–ľâ–ş
â—„vae_nameâ–ľâ–ş
â—„clip_nameâ–ľâ–ş
â—„typeâ–ľâ–ş
â—„deviceâ–ľâ–ş
Categoryđź’€Prepack
Inputs (6)
| Name | Type | Default | Description |
|---|---|---|---|
| unet_name | COMBO | The name of the diffusion model (UNET) to load. | |
| weight_dtype | COMBO | The weight dtype for the diffusion model. | |
| vae_name | COMBO | The VAE to load. | |
| clip_name | COMBO | The name of the CLIP model to load. | |
| type | COMBO | The type of CLIP configuration to use. | |
| deviceopt | COMBO | Device override for CLIP loading/offloading. Use 'cpu' to force CPU; 'default' lets Comfy manage devices. |
Outputs (3)
| Name | Type | Description |
|---|---|---|
| model | MODEL | The loaded diffusion model (UNET) used for denoising latents. |
| clip | CLIP | The loaded CLIP model used for encoding text prompts. |
| vae | VAE | The loaded VAE for encoding/decoding latents (can be None if loading fails). |