floyo logo
Powered by
ThinkDiffusion
floyo logo
Powered by
ThinkDiffusion

Inpainting with reference image

66

Ever needed to add something to an image that wasn't there before? But your new objects has a visual reference? Try this workflow that uses Flux fill and Flux redux.

Upload a base image and a reference image (your object).
Mask your base image and run it.
No prompt necessary.

Local install
You'll need to download these 6 recommended models and 1 model that will be auto-downloaded.

- sigclip_vision_patch14_384.safetensors .../comfyui/models/clip_vision/
- flux1-fill-dev.safetensors .../comfyui/models/diffusion_models/
- flux1-redux-dev.safetensors .../comfyui/models/style_models/
- t5xxl_fp16.safetensors .../comfyui/models/clip/
- clip_l.safetensors .../comfyui/models/clip/
- gokaygokay/Florence-2-SD3-Captioner (auto-download)
- ae.safetensors .../comfyui/models/vae/

Models can be downloaded from Comfy.Org or the model manager.

Read more

N
Generates in about -- secs

Nodes & Models

CLIPVisionLoader
sigclip_vision_patch14_384.safetensors
DualCLIPLoader
t5xxl_fp16.safetensors
clip_l.safetensors
UNETLoader
flux1-fill-dev.safetensors
StyleModelLoader
flux1-redux-dev.safetensors
Label (rgthree)
VAELoader
ae.safetensors
Note
LoadImage
CLIPTextEncode
DifferentialDiffusion
MaskToImage
PreviewImage
CLIPVisionEncode
ImageScale
ImageToMask
InpaintModelConditioning
StyleModelApply
FluxGuidance
KSampler
VAEDecode
ImageCrop
SaveImage
DownloadAndLoadFlorence2Model
Florence2Run
DownloadAndLoadFlorence2Model
Florence2Run
Text Multiline
Text List
Text List to Text
ImageResizeKJ
ImageResizeKJ
ImageConcanate
ImageResizeKJ
ImageConcanate
ImageGenResolutionFromImage
ImageGenResolutionFromImage
ImageGenResolutionFromImage
CR Color Panel
ShowText|pysssss
ShowText|pysssss

Ever needed to add something to an image that wasn't there before? But your new objects has a visual reference? Try this workflow that uses Flux fill and Flux redux.

Upload a base image and a reference image (your object).
Mask your base image and run it.
No prompt necessary.

Local install
You'll need to download these 6 recommended models and 1 model that will be auto-downloaded.

- sigclip_vision_patch14_384.safetensors .../comfyui/models/clip_vision/
- flux1-fill-dev.safetensors .../comfyui/models/diffusion_models/
- flux1-redux-dev.safetensors .../comfyui/models/style_models/
- t5xxl_fp16.safetensors .../comfyui/models/clip/
- clip_l.safetensors .../comfyui/models/clip/
- gokaygokay/Florence-2-SD3-Captioner (auto-download)
- ae.safetensors .../comfyui/models/vae/

Models can be downloaded from Comfy.Org or the model manager.

Read more

N