ComfyUI Nodes Reference

How ComfyUI Nodes Reference works on Windows

Core loaders, KSampler nodes, ControlNet preprocessors, IPAdapter, and ComfyUI Manager custom node packs on Windows. Wiring latent inputs, conditioning links, VAE decode steps, and custom node packages without breaking your Python environment.

Every page in this directory focuses on practical local execution on Windows 10 and Windows 11, keeping your checkpoints, LoRAs, and generated outputs entirely on your own SSD.

Start by grabbing the current Windows build from /download, or browse the guides and node blueprints linked below.

Key workflow principles

• Keep large `.safetensors` and `.gguf` weights inside organized `models/` subfolders or map shared directories through extra_model_paths.yaml.
• Match your model precision (FP16, FP8 e4m3fn, or GGUF Q4/Q8) to your card's VRAM tier so sampling stays in GPU memory without paging to disk.
• Drag any generated PNG directly onto the ComfyUI Studio canvas to reload its exact node wiring, seed, and sampler values.

Featured pages in ComfyUI Nodes Reference

• KSampler & KSampler (Advanced) — The heart of every ComfyUI graph: iteratively denoises a latent tensor using your chosen solver, scheduler curve, and classifier-free guidance scale. Package: Built-in Core.
• Wiring & Fixing KSampler & KSampler (Advanced) — Practical pin-by-pin connection order and common error fixes for KSampler & KSampler (Advanced) in ComfyUI workflows.
• Load Checkpoint (CheckpointLoaderSimple) — Loads unified safetensors checkpoints from models/checkpoints/ and exposes the UNet model, text encoder, and VAE output pins. Package: Built-in Core.
• Wiring & Fixing Load Checkpoint (CheckpointLoaderSimple) — Practical pin-by-pin connection order and common error fixes for Load Checkpoint (CheckpointLoaderSimple) in ComfyUI workflows.
• CLIP Text Encode (Prompt) — Tokenizes positive and negative text prompts, supporting (word:1.2) attention weights and embedding: tags. Package: Built-in Core.
• Wiring & Fixing CLIP Text Encode (Prompt) — Practical pin-by-pin connection order and common error fixes for CLIP Text Encode (Prompt) in ComfyUI workflows.
• VAE Decode & VAE Encode (Tiled) — Translates compressed 8x latent tensors into RGB images and back, with tiled variants that eliminate out-of-memory spikes at 4K and in video runs. Package: Built-in Core.
• Wiring & Fixing VAE Decode & VAE Encode (Tiled) — Practical pin-by-pin connection order and common error fixes for VAE Decode & VAE Encode (Tiled) in ComfyUI workflows.
• Apply ControlNet (Advanced) — Injects spatial structure from depth maps, OpenPose skeletons, or Canny edges across a chosen percentage window of sampling steps. Package: Built-in Core.
• Wiring & Fixing Apply ControlNet (Advanced) — Practical pin-by-pin connection order and common error fixes for Apply ControlNet (Advanced) in ComfyUI workflows.
• Load LoRA (LoraLoader) — Chains low-rank adaptation weights onto the UNet and CLIP streams with independent model and text-encoder multipliers. Package: Built-in Core.
• Wiring & Fixing Load LoRA (LoraLoader) — Practical pin-by-pin connection order and common error fixes for Load LoRA (LoraLoader) in ComfyUI workflows.
• ComfyUI-Manager — Essential management extension that detects missing red nodes in imported workflows and installs custom node repositories with one click. Package: ltdrdata/ComfyUI-Manager.
• Wiring & Fixing ComfyUI-Manager — Practical pin-by-pin connection order and common error fixes for ComfyUI-Manager in ComfyUI workflows.