comfyui-custom-node

High quality training free inpaint for every stable diffusion model. Supports ComfyUI

Generate, downscale, change palletes and restore pixel art images with SDXL. Nodes for Astropulse/pixeldetector, Palette Loader from Image, Grid view for palettes etc.

Refine and improve the quality of small faces in MiniMax H3 video. Per-frame face tracking, crop, refine with H3, stitch back.

MCP server + embedded terminal that lets Claude Code see and edit your ComfyUI workflows

TBG Enhanced Tiled Upscaler and Refiner upscales up to 200MP with precise control. It features dual-model processing (structure + detail), per-pixel denoising to control where creativity is applied, and per-tile and per-object prompts, denoise, and seeds. Advanced fast previews. New generative tile fusion seamless joining, even on high-denoise.

Executable Stable Diffusion merge recipes in comfyui

The local-first sidebar AI agent for ComfyUI — runs on your own Claude OR ChatGPT subscription (no API keys, no extra LLM costs). Drives your live graph: edits, workflows & installs in natural language. The panel UI for comfyui-mcp, the agent-native control plane for ComfyUI.

Nodes to run Hunyuan Image 3 locally with BF16 and NF4 quantized options in Comfyui

ComfyUI custom nodes for Diffusion Attentive Attribution Maps (DAAM)

Use your own API keys with ComfyUI. No account required.

Automatically benchmark and optimize attention in diffusion models. 1.5-2x speedup on RTX 4090.

Local LLM chat panel for ComfyUI — LM Studio & Ollama, multi-session, vision support, Guide Materials

ComfyUI extension for automatically resolving missing models in loaded workflows. It features intelligent local fuzzy matching, direct cloud downloads, background download tracking, and automated in-place workflow updating.

Detect and resolve missing models in ComfyUI workflows — active-graph analysis, multi-source resolution (Civitai / Hugging Face), and checksum-verified downloads.

ComfyUI nodes useful to Tweak SDXL, Illustrious, and NAI models by adjusting specific 'slices'. These nodes break the U-Net and CLIP into areas of application (like Style, Lighting, Word Logic...) and use sliders to boost or lower the intensity of any of those sections without training.