ComfyUI Node:  JLC CaptionForge Pipeline Planner

Authored by Damkohler

Created

Updated

1 stars

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

Category

Captioning/CaptionForge

Inputs

Planner - enabled BOOLEAN
Input - image path STRING
Input - recursive BOOLEAN
Input - filename glob STRING
Output - folder STRING
Output - run name STRING
Output - overwrite outputs BOOLEAN
LoRA - trigger word STRING
LoRA - user caption anchor STRING
Caption - Joy runs/image
  • Disabled
  • 1
  • 2
  • 3
  • 4
  • 5
Caption - Qwen runs/image
  • Disabled
  • 1
  • 2
  • 3
  • 4
  • 5
Caption - Ollama runs/image
  • Disabled
  • 1
  • 2
  • 3
  • 4
  • 5
Caption - base seed INT
Caption - seed mode
  • fixed
  • increment
  • decrement
  • random
Caption - temperature schedule STRING
Caption - top p schedule STRING
Caption - top k schedule STRING
Caption - max image size INT
Caption - max new tokens INT
Distiller - model
  • mistral-small:24b
  • VladimirGav/gemma4-26b-16GB-VRAM-Uncensored
  • deepseek-r1:32b
  • tarruda/neuraldaredevil-8b-abliterated:fp16
  • gpt-oss:20b
  • custom
Distiller - custom Ollama model STRING
Distiller - base seed INT
Distiller - seed mode
  • fixed
  • increment
  • decrement
  • random
Distiller - strategy
  • single_pass
  • by_model_then_global
Distiller - max caption chars for LLM INT
Distiller - num predict INT
Distiller - temperature FLOAT
Distiller - top p FLOAT
Distiller - top k INT
Distiller - write prompt JSONL BOOLEAN
Distiller - preserve raw response BOOLEAN
Validator - model
  • gemma4:26b
  • qwen3.6:35B-A3B
  • huihui_ai/gemma-4-abliterated:26b
  • custom
Validator - custom Ollama model STRING
Validator - base seed INT
Validator - seed mode
  • fixed
  • increment
  • decrement
  • random
Validator - num predict INT
Validator - temperature FLOAT
Validator - top p FLOAT
Validator - top k INT
Validator - write prompt JSONL BOOLEAN
Validator - preserve raw VLM response BOOLEAN
Final - caption style
  • narrative
  • comma
  • both
Final - write TXT sidecars BOOLEAN
Final - write JSONL BOOLEAN
Input - single image IMAGE

Outputs

IMAGE

CAPTIONFORGE_PIPELINE_PLAN

STRING

Extension: CaptionForge

A model-agnostic captioning framework for ComfyUI designed to generate cleaner, more consistent, auditable captions for LoRA dataset preparation through structured comparison, repeated observation, and consensus-based refinement.

Authored by Damkohler

View Nodes

Looking for a different node?

More nodes in CaptionForge

Run ComfyUI workflows without the setup

No installs, no CUDA version roulette, no GPU sitting idle on your bill. Bring a workflow and run it in the browser.

Learn more