SeaArt ComfyUI WIKI
  • SeaArt ComfyUI WIKI
  • Core Nodes
    • Advanced
      • conditioning
        • CLIP Text Encode SDXL
        • CLIP Text Encode SDXL Refiner
        • Conditioning Set Timestep Range
        • Conditioning Zero Out
        • Flux
          • CLIPTextEncodeFlux Node for ComfyUI Explained
          • FluxGuidance - ComfyUI Node Functionality Description
      • Loaders
        • Load CLIP
        • Load Checkpoint With Config (DEPRECATED)
        • Diffusers Loader
        • Dual CLIP Loader - How it work and how to use it
        • UNET Loader Guide | Load Diffusion Model
      • Model
        • Model Sampling Continuous EDM
        • Model Sampling Discrete
        • Rescale CFG
      • Model-merging
        • Checkpoint Save
        • CLIPMerge Simple
        • CLIP Save
        • Model Merge Add
        • Model Merge Blocks
        • Model Merge Simple
        • Model Merge Subtract
        • VAE Save
    • Conditioning
      • 3d-models
        • Stable Zero 123 Conditioning
        • Stable Zero 123 Conditioning Batched
      • CLIP Set Last Layer
      • CLIP Text Encode (Prompt)
      • CLIP Vision Encode
      • Conditioning Average
      • Conditioning (Combine)
      • Conditioning (Concat)
      • Conditioning (Set Area)
      • Conditioning (Set Area with Percentage)
      • Conditioning Set Area Strength
      • Conditioning (Set Mask)
      • Apply ControlNet
      • Apply ControlNet (Advanced)
      • gligen
        • GLIGEN Text Box Apply
      • inpaint
        • Inpaint Model Conditioning
      • style-model
        • Apply Style Model
      • unCLIP Conditioning
      • upscale-diffusion
        • SD_4X Upscale Conditioning
      • video-models
        • SVD img2vid Conditioning
    • Image
      • animation
        • Save Animated PNG
        • Save Animated WEBP
      • batch
        • Image From Batch
        • Rebatch Images
        • Repeat Image Batch
      • Empty Image
      • Batch Images
      • Image Composite Masked
      • Invert Image
      • Pad Image for Outpainting
      • Load Image
      • postprocessing
        • Image Blend
        • Image Blur
        • Image Quantize
        • Image Sharpen
      • preprocessors
        • Canny Node
      • Preview Image
      • Save Image
      • transform
        • Image Crop
      • upscaling
        • Upscale Image
        • Upscale Image By
        • Image Scale To Total Pixels
        • Upscale Image (using Model)
    • Latent
      • Latent Upscale
      • Empty Latent Image
      • Latent Upscale By
      • Latent Composite
      • VAE Decode
      • VAE Encode
      • Latent Composite Masked
      • advanced
        • Latent Add
        • Latent Batch Seed Behavior
        • Latent Interpolate
        • Latent Multiply
        • Latent Subtract
      • batch
        • Latent Batch
        • Latent From Batch
        • Rebatch Latents
        • Repeat Latent Batch
      • inpaint
        • Set Latent Noise Mask
        • VAE Encode (for Inpainting)
      • transform
        • Crop Latent
        • Flip Latent
        • Rotate Latent
    • Loaders
      • Checkpoint Loader (Simple)
      • CLIP Vision Loader
      • ControlNet Loader
      • Diff ControlNet Loader
      • GLIGEN Loader
      • Hypernetwork Loader
      • Lora Loader
      • Lora Loader Model Only
      • Style Model Loader
      • unCLIP Checkpoint Loader
      • Upscale Model Loader
      • VAE Loader
      • video-models
        • Image Only Checkpoint Loader (img2vid model)
    • Mask
      • compositing
        • Join Image with Alpha
        • Porter-Duff Image Composite
        • Split Image with Alpha
      • Crop Mask
      • Feather Mask
      • Grow Mask
      • Image Color To Mask
      • Image To Mask
      • Invert Mask
      • Load Image (as Mask)
      • Mask Composite
      • Mask To Image
      • Solid Mask
    • Sampling
      • custom-sampling
        • SamplerCustom
        • samplers
          • KSampler Select
          • Sampler DPMPP_2M_SDE
          • Sampler DPMPP_SDE
        • schedulers
          • Basic Scheduler
          • Exponential Scheduler
          • Karras Scheduler
          • Polyexponential Scheduler
          • SD Turbo Scheduler
          • VP Scheduler
        • sigmas
          • Flip Sigmas
          • Split Sigmas
      • KSampler
      • KSampler (Advanced)
      • Sampler
      • video-models
        • Video Linear CFG Guidance
    • Utils
      • Note
      • Primitive
      • Reroute
      • Terminal Log (Manager)
  • ComfyUI Workflow Example
    • 1-Img2Img
    • 2-2 Pass Txt2Img
    • 3-Inpaint
    • 4-Area Composition
    • 5-Upscale Models
    • 6-LoRA
    • 7-ControlNet
    • 8-Noisy Latent Composition
    • 9-Textual Inversion Embeddings
    • 10-Edit Models
    • 11-Model Merging
    • 12-SDXL
    • 13-Stable Cascade
    • 14-UnCLIP
    • 15-Hypernetworks
    • 16-Gligen
    • 17-3D Examples
    • 18-Video
    • 19-LCM Examples
    • 20-ComfyUI SDXL Turbo Examples
  • How to publish as an AI app
  • FAQ (Must see!!!)
Powered by GitBook
On this page
  • Documentation
  • Input types
  • Output types
  1. Core Nodes
  2. Advanced
  3. Model-merging

CLIPMerge Simple

PreviousCheckpoint SaveNextCLIP Save

Last updated 5 months ago

Documentation

  • Class name: CLIPMergeSimple

  • Category: advanced/model_merging

  • Output node: False

This node specializes in merging two CLIP models based on a specified ratio, effectively blending their characteristics. It selectively applies patches from one model to another, excluding specific components like position IDs and logit scale, to create a hybrid model that combines features from both source models.

Input types

Parameter
Comfy dtype
Description

clip1

CLIP

The first CLIP model to be merged. It serves as the base model for the merging process.

clip2

CLIP

The second CLIP model to be merged. Its key patches, except for position IDs and logit scale, are applied to the first model based on the specified ratio.

ratio

FLOAT

Determines the proportion of features from the second model to blend into the first model. A ratio of 1.0 means fully adopting the second model’s features, while 0.0 retains only the first model’s features.

Output types

Parameter
Comfy dtype
Description

clip

CLIP

The resulting merged CLIP model, incorporating features from both input models according to the specified ratio.