Skip to content

⬡ KSampler

Central hub node: receives models, settings, prompts, and images — samples and produces the generation pipeline.

Inputs

Name Type Required Description
model_bundle UME_BUNDLE Model + CLIP + VAE from a Loader node
settings UME_SETTINGS Parameters from Generation Settings
positive POSITIVE Positive prompt text
negative NEGATIVE Negative prompt text
loras UME_LORA_STACK LoRA stack from LoRA Block nodes
image UME_IMAGE Image bundle for img2img/inpaint/outpaint

Outputs

Name Type Description
gen_pipe UME_PIPELINE Complete pipeline with generated image, ready for post-processing or saving

What Happens Inside

The KSampler orchestrates the full generation pipeline:

  1. Validates the model bundle (model, clip, vae must all be present)
  2. Applies LoRAs from the stack (if connected)
  3. Encodes prompts via CLIP (with caching for repeated prompts)
  4. Applies ControlNets from image bundle (if present)
  5. Prepares latent — empty for txt2img, or VAE-encoded for img2img/inpaint
  6. Samples using the configured sampler + scheduler + steps
  7. Decodes the latent to pixel space via VAE
  8. Packs everything into a UME_PIPELINE for downstream nodes

Optimization

The KSampler caches prompt encodings and ControlNet models. If you change only the seed, re-encoding is skipped for faster iteration.