ComfyUI Node
TextEncodeReferenceImages
Encodes text with reference images for VL models and conditioning.
TextEncodeReferenceImages
- clip
- vae
- image_encoder
- images
- CONDITIONING
◄prompt—►
◄vl_selectionnone►
◄auto_resize_imagestrue►
Categoryconditioning
Inputs (7)
| Name | Type | Default | Description |
|---|---|---|---|
| clip | CLIP | The CLIP model to use for encoding. | |
| prompt | STRING | The text prompt to encode. | |
| vl_selection | COMBO | none | Select the Vision-Language model strategy. |
| auto_resize_images | BOOLEAN | true | Automatically resize reference images to 1024x1024. |
| vaeopt | VAE | VAE for encoding reference images into latents. | |
| image_encoderopt | CLIP_VISION | CLIP Vision model for encoding reference images. | |
| imagesopt | IMAGE | Reference images to be used for conditioning (supports image batches). |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| CONDITIONING | CONDITIONING | — |