floyo logo
Powered by
ThinkDiffusion
Pricing
Wan 2.7 is now live. Check it out ๐Ÿ‘‰๐Ÿผ
floyo logo
Powered by
ThinkDiffusion
Pricing
Wan 2.7 is now live. Check it out ๐Ÿ‘‰๐Ÿผ

Character Consistency - 1 image only. Flux Ace++

Load 1 image and use natural language to get variations of your input image. Works great for character consistency to generate a dataset for loras.

834

Generates in about 42 secs

Nodes & Models

VAELoader
ae.safetensors
LoadImage
Label (rgthree)
UNETLoader
flux1-fill-dev.safetensors
DualCLIPLoader
clip_l.safetensors
t5xxl_fp16.safetensors
Note
CLIPTextEncode
ImagePadForOutpaint
DifferentialDiffusion
FluxGuidance
InpaintModelConditioning
KSampler
VAEDecode
ImageCrop
PreviewImage
SaveImage
ImageResize+
ImageAndMaskPreview
ImageAndMaskPreview

Character Consistency with Flux Ace++

Only 1 image required for consistency

Load 1 image and use natural language to get variations of your input image. Works great for character consistency to generate a dataset for loras.

Load an image in the load image node. The workflow automatically adjusts the size.
Example prompt: Maintain the face and hat. Man in a red shirt is having a coffee in a coffeeshop.
Example prompt: Maintain the character, seen from side view.

Run times

Cold start: 2 minutes
Hot start: 30 seconds

Read more

N