filliptm/ComfyUI_Fill-Nodes

What it solves

Fill-Nodes is a comprehensive collection of custom nodes for ComfyUI designed to bridge the gap between simple image generation and complex production workflows. It provides a wide array of tools for advanced image manipulation, visual effects (VFX), file management, and integration with external AI services, reducing the need for multiple disparate plugins.

How it works

It extends ComfyUI by adding specialized nodes across several categories:

• Image & Video Processing: Tools for line extraction, aspect ratio cropping, batching, and video export (WebM/WebP).
• VFX & Shaders: Implements digital effects like pixel sorting, glitching, ASCII art, and GLSL shader rendering via Shadertoy.
• AI Integration: Provides interfaces to connect workflows to external models and APIs including GPT, DALL-E, Gemini, Hugging Face, and Fal.ai.
• Data & File Handling: Nodes for PDF creation, CSV extraction/saving, and Google Drive integration.
• Captioning: Tools for generating captions (via Ollama), saving them to disk, and analyzing word frequency.

Who it’s for

It is built for ComfyUI users who need professional-grade utility tools for batch processing, dataset preparation (captioning/cropping), and creative visual experimentation within their generative AI pipelines.

Highlights

• Advanced VFX: Includes depth-of-field blur, halftone patterns, and infinite zoom effects.
• Production Utilities: Features an interactive image picker to pause workflows for manual selection and a padding remover for cleaning datasets.
• External AI Connectivity: Direct nodes for interacting with various LLMs and image generation APIs.
• Dataset Tools: Comprehensive CSV and PDF tools for organizing and documenting image-caption pairs.