ComfyUI Node
Unload Models By Type
Unload currently resident ComfyUI models by type. Use it inline as a passthrough before CLIP Text Encode to unload the diffusion model, or after CLIP Text Encode to unload the text encoder before sampling.
Unload Models By Type
- trigger
- model
- clip
- vae
- conditioning
- trigger
- model
- clip
- vae
- conditioning
- status
◄model_typeText Encoder►
◄device_scopeCurrent Device►
◄empty_cachetrue►
CategoryWorkflowX/VRAM
Inputs (8)
| Name | Type | Default | Description |
|---|---|---|---|
| model_type | COMBO | Text Encoder | 6 options: Text Encoder, Diffusion Model / UNet, VAE, CLIP Vision, Other Loaded Models, All Loaded Models |
| device_scope | COMBO | Current Device | 2 options: Current Device, All Devices |
| empty_cache | BOOLEAN | true | — |
| triggeropt | * | — | |
| modelopt | MODEL | — | |
| clipopt | CLIP | — | |
| vaeopt | VAE | — | |
| conditioningopt | CONDITIONING | — |
Outputs (6)
| Name | Type | Description |
|---|---|---|
| trigger | * | — |
| model | MODEL | — |
| clip | CLIP | — |
| vae | VAE | — |
| conditioning | CONDITIONING | — |
| status | STRING | — |