ComfyUI Documentation

Official docs for ComfyUI — the most powerful and modular Stable Diffusion GUI and backend.

CRITICAL: grep references/ for keywords before answering.

Reference Index (317 docs)

Built-in Nodes (67 docs)

• references/built-in-nodes/ — Every built-in node with inputs, outputs, parameters:
  • Loaders: CheckpointLoaderSimple, ClipLoader, ClipSave, LoraLoader, VAELoader, UNETLoader, ControlNetLoader, StyleModelLoader
  • Conditioning: ClipTextEncode, ClipTextEncodeFlux, ConditioningCombine, ConditioningSetArea, ConditioningSetMask, GLIGEN
  • Sampling: KSampler, KSamplerAdvanced, BasicScheduler, KarrasScheduler, SamplerCustom
  • Latent: EmptyLatentImage, LatentUpscale, LatentComposite, LatentBlend, LatentCrop
  • Image: LoadImage, SaveImage, PreviewImage, ImageScale, ImageInvert, ImageBatch, ImagePadForOutpaint
  • Mask: MaskComposite, SetLatentNoiseMask, ImageToMask
  • ControlNet: Canny, ControlNetApply, ControlNetApplyAdvanced
  • Video/Audio: LoadVideo, SaveVideo, LoadAudio, SaveAudio
  • Advanced: ModelMerge, DifferentialDiffusion, FluxGuidance, InstructPixToPixConditioning