Have a ComfyUI workflow? Get a readable Workflow Doc in your browser.Make a workflow readable.Try it now ↗
This workflow uses the Flux Upscale Controlnet ( ) to upscale a provided image. Florence 2 is used to caption the image for automatic prompting.
Runtime profile
Source description
This workflow uses the Flux Upscale Controlnet ( https://huggingface.co/jasperai/Flux.1-dev-Controlnet-Upscaler ) to upscale a provided image. Florence 2 is used to caption the image for automatic prompting.
A script by Kijai ( https://huggingface.co/Kijai/flux-fp8/discussions/7#66ae0455a20def3de3c6d476 ) is used to convert the controlnet to fp8 so that Flux (fp8) and the controlnet can fit into 16GB of VRAM.
Important Notes for the Workflow
AI-generated commentary
AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.
Flux Controlnet Upscale is a Flux.1 D workflow for upscaling a provided image: it uses Flux Upscale Controlnet and Florence 2, which captions the image for automatic prompting, and produces an upscaled image.
Input: a provided image. Output: an upscaled image.
The output image size follows the input image’s aspect ratio and the value in the Scale To Megapixels node.
Install Flux VAE with the manager before use.
Suggestion · not verified
The FP8 ControlNet conversion is intended to let Flux FP8 and the ControlNet fit within 16 GB of VRAM.
Named model files include ae.safetensors; flux1-dev-fp8-e4m3fb.safetensors; flux1-dev-fp8-e4m3fn.safetensors; Flux.1-dev-Controlnet-Upscaler_fp8.safetensors; t5xxl_fp8_e4m3fn_scaled.safetensors; and ViT-L-14-BEST-smooth-GmP-TE-only-HF-format.safetensors.
Other named models are Flux.1-dev-Controlnet-Upscaler, Flux VAE, and microsoft/Florence-2-base.
Named workflow nodes include CLIPTextEncode; ColorMatch; ConditioningZeroOut; ControlNetApplyAdvanced; ControlNetLoader; Display Any (rgthree); DownloadAndLoadFlorence2Model; DualCLIPLoader; EmptyLatentImage; Florence2Run; FluxGuidance; GetImageSizeAndCount; ImageScaleToMegapixel
Set Scale To Megapixels between 1 and 2; output size follows this setting and the input image’s aspect ratio.
Suggestion · not verified
Set ControlNet strength between 0.5 and 0.6.
Suggestion · not verified
Use end_percent to adjust fidelity: higher values keep the upscale closer to the original, while lower values leave more room for model creativity.
Suggestion · not verified
For an image that is not a photo, bypass the Text Find and Replace node.
Suggestion · not verified
Does this need to be edited?
Sign in to send an edit request.
Sources
1 sourceSource excerpts
1 excerptSource context: 4330 downloads · Type Workflows · Base model Flux.1 D
You should bypass the "Text Find and Replace" node if you are not upscaling a photo.
The size of the output image is controlled by the aspect ratio of the input image and the number of megapixels in the "Scale To Megapixels" node. Set this value between 1 - 2 for the best results.
Strength of the controlnet should be between 0.5 - 0.6
The end_percent can be used to control how faithful the upscale is to the provided image. Higher values to be more faithful to the original image and lower values to give the model more room for creativity.
axior from reddit reported good results with the flux + upscale controlnet method for upscaling to 8megapixels, Controlnet strength 0.85, up until 0.85 of generation, denoise 0.95
Models
flux1-dev-fp8-e4m3fb.safetensors (models/diffusion_models/flux): https://huggingface.co/Kijai/flux-fp8/tree/main
t5xxl_fp8_e4m3fn_scaled.safetensors (models/clip): https://huggingface.co/comfyanonymous/flux_text_encoders/tree/main
ViT-L-14-BEST-smooth-GmP-TE-only-HF-format.safetensors (models/clip): https://huggingface.co/zer0int/CLIP-GmP-ViT-L-14/tree/main
Flux.1-dev-Controlnet-Upscaler (models/controlnet): https://huggingface.co/jasperai/Flux.1-dev-Controlnet-Upscaler converted to fp8 using script by Kijai ( https://huggingface.co/Kijai/flux-fp8/discussions/7#66ae0455a20def3de3c6d476 ) for 16GB VRAM users.
Flux VAE: install with manager
Custom Nodes
ComfyUI-Florence2 by Kijai
WAS Node Suite
comfyui-art-venture
rgthree's ComfyUI Nodes
KJNodes for ComfyUI
Estimated VRAM requirement
Estimate unavailable
320 MB across 1 of 9 model files. Model file total + 25% loading overhead + 2 GB execution buffer, rounded up.
Requirements
27 requirementsae.safetensors
FoundVAE · 320 MB · SAFETENSORS · Unknown
DualCLIPLoader
PossibleText encoder · Unknown
Flux.1-dev-Controlnet-Upscaler_fp8.safetensors
Not resolvedUpscaler · Unknown
flux1-dev-fp8-e4m3fn.safetensors
File unverifiedCheckpoint · Unknown · SHA-256 47d8dbdc6da6
Vision model · Hugging Face · microsoft/Florence-2-base
t5xxl_fp8_e4m3fn_scaled.safetensors
File unverifiedText encoder · Unknown
VAEDecode
PossibleVAE · Unknown
VAELoader
PossibleVAE · Unknown
ViT-L-14-BEST-smooth-GmP-TE-only-HF-format.safetensors
Not resolvedText encoder · Unknown
CLIPTextEncode
PossibleNode pack · Unknown
ColorMatch
PossibleNode pack · Unknown
ConditioningZeroOut
PossibleNode pack · Unknown
ControlNetApplyAdvanced
Node pack · Unknown
ControlNetLoader
PossibleNode pack · Unknown
Display Any (rgthree)
Not resolvedNode pack · Unknown
DownloadAndLoadFlorence2Model
PossibleNode pack · Unknown
EmptyLatentImage
PossibleNode pack · Unknown
Florence2Run
PossibleNode pack · Unknown
FluxGuidance
Not resolvedNode pack · Unknown
GetImageSizeAndCount
PossibleNode pack · Unknown
Node pack · Unknown
KSampler
PossibleNode pack · Unknown
LoadImage
PossibleNode pack · Unknown
SaveImage
PossibleNode pack · Unknown
Node pack · Unknown
TorchCompileModel
Not resolvedNode pack · Unknown
UNETLoader
PossibleNode pack · Unknown
