Nodes/OllamaTools for ComfyUI/🖼️ Pic Describer
ComfyUI Node

🖼️ Pic Describer

A ComfyUI node in OllamaTools with 5 inputs and 1 output.

By max-dingsda·Created about a year ago·Updated 5 months ago· 4
🖼️ Pic Describer
  • image
  • description
modelministral-3:latest
styleneutral
promptdescribe the image
cleanup_outputtrue
CategoryOllamaTools

Inputs (5)

NameTypeDefaultDescription
imageIMAGEThe image you want the LLM to describe. Usually from 'Load Image'.
modelCOMBOministral-3:latestMultimodal LLM used for describing the image.
styleCOMBOneutralChoose how the image should be interpreted: factually, creatively, or filtered for SFW/NSFW.
promptSTRINGdescribe the imageAdditional instruction or focus for the image description. Used alongside style.
cleanup_outputBOOLEANtrueRemove hallucinations and verbose reasoning from the LLM output.

Outputs (1)

NameTypeDescription
descriptionSTRING