ComfyUI Node
Krea2 Edit (source patch)
Adds the krea2_edit in-context source-preservation path (source latent as frame=1 tokens) to a Krea2 model.
Krea2 Edit (source patch)
- model
- source_latent
- source_latent_b
- ref_boost_mask
- vae
- source_image
- source_image_b
- target_latent
- MODEL
◄ref_boost1.00►
◄ref_boost_a1.00►
◄fit_modefit►
Categorykrea2edit
Inputs (11)
| Name | Type | Default | Description |
|---|---|---|---|
| model | MODEL | — | |
| source_latent | LATENT | — | |
| source_latent_bopt | LATENT | 2nd reference (subject photo) for multi-ref LoRAs -> RoPE frame=2, training-matched order: scene first, subject second | |
| ref_boostopt | FLOAT | 1.000–1000 | reference-fidelity dial: multiplies target->reference attention. Applies to the LAST ref (= the subject in two-ref workflows, the only ref in single-ref). 1.0 = off, >1 pulls harder toward the reference's appearance, <1 loosens. Optimal value is model-specific |
| ref_boost_aopt | FLOAT | 1.000–1000 | same dial for the FIRST ref (= the scene in two-ref workflows). No effect in single-ref workflows. 1.0 = off |
| fit_modeopt | COMBO | fit | how an image source fits a mismatched output aspect ratio (needs vae + source_image connected): fit = resample the source to the target grid at a centered offset — matches how this model was trained (default, use this); crop (legacy) = center-crop to the target AR then resize (v1/v1.1 geometry, only for older weights) |
| ref_boost_maskopt | MASK | optional region on the (last) reference to boost, e.g. the face; empty = whole reference | |
| vaeopt | VAE | RECOMMENDED with source_image: enables the blur-proof pixel-space path (crop+resize in pixels, encode internally) — immune to input/output resolution mismatches | |
| source_imageopt | IMAGE | source as IMAGE (with vae connected): overrides source_latent with exact pixel-space fitting — fixes blurry results from mismatched resolutions | |
| source_image_bopt | IMAGE | 2nd reference as IMAGE (with vae) | |
| target_latentopt | LATENT | RECOMMENDED with vae + source_image: wire the SAME latent you feed KSampler.latent_image. Lets the node VAE-encode the source here, before sampling starts, instead of on the first step — otherwise the VAE is pulled onto the GPU mid-sampling and can evict part of the diffusion model, slowing every remaining step on VRAM-tight setups |
Outputs (1)
| Name | Type | Description |
|---|---|---|
| MODEL | MODEL | — |