EasyUse/Imagegenerated

Image To Prompt(imageInterrogator)

The Image To Prompt node interrogates an input image and returns a text prompt that describes it.

image Interrogator

image
prompt
mode
COMBO
use_lowvram
Easy Use

Description

The Image To Prompt node interrogates an input image and returns a text prompt that describes it. It is part of the EasyUse/Image category and is designed to bridge image inputs and prompt-based generation workflows.

Inputs

NameTypeDefaultDescription
imageIMAGEThe input image to interrogate and convert into a text prompt.
modeCOMBOThe interrogation mode to use. Options are fast, classic, best, and negative.
use_lowvramBOOLEANtrueWhen enabled, uses a low VRAM mode to reduce GPU memory usage during interrogation.

Outputs

TypeDescription
STRINGThe generated text prompt.

Usage Notes

  • mode controls the type and quality of the generated prompt:

    • fast — Quick interrogation, producing a shorter or less detailed prompt.
    • classic — Standard interrogation behavior.
    • best — Higher-quality or more detailed prompt, typically with slower processing.
    • negative — Generates a negative prompt rather than a positive description.
  • use_lowvram is enabled by default. Leave it enabled on systems with limited GPU memory; disable it only if you have enough VRAM and want to avoid any performance overhead from low VRAM handling.

  • The output is a plain string and can be connected directly to nodes that accept STRING prompt input, such as conditioning text encoders.

Comments

Sign in with GitHub to join the discussion.

Loading comments…