Optimization of the interface and addition of new options.
Runtime profile
Requirements
Upscaler · 63.9 MB · Hugging Face · gemasai/4x_NMKD-Siax_200k
VAE · 320 MB · SAFETENSORS · Hugging Face · UmeAiRT/ComfyUI-Auto_installer
AI-generated commentary
AI-generated explanation based on source and configuration details. Suggestions are clearly labeled.
Use 【FLUX】IMG to IMG — v4.4 in ComfyUI to turn a source image, written prompt, and loaded FLUX model into a generated image with optional SeedVR2 upscaling and embedded generation parameters.
This eight-node ComfyUI workflow uses the UmeAiRT Toolkit for image-to-image generation, optional SeedVR2 tiled upscaling, embedded generation parameters, and three LoRA slots.
Provide a source image and written prompt, then load a FLUX model; three LoRA slots are also available.
Output images are saved with generation parameters embedded for online publishing and remixing.
In ComfyUI, enable “Use Nodes V2 (Vue)” in Settings, refresh the page, and then load the workflow; leaving it off may cause display problems.
Install ComfyUI-UmeAiRT-Toolkit through ComfyUI Manager by searching “UmeAiRT”, or use the UmeAiRT Auto-Installer.
Listed model files include 4x_NMKD-Siax_200k.pth, ae.safetensors, clip_l.safetensors, flux1-dev-Q8_0.gguf, flux1-dev-fp8.safetensors, flux1-dev-fp16.safetensors, t5-v1_1-xxl-encoder-Q8_0.gguf, t5xxl_fp8_e4m3fn.safetensors, and ViT-L-14-TEXT-detail-improved-hiT-GmP-TE-only-HF.sภet
Node names include BasicScheduler, DualCLIPLoaderGGUF, Fast Groups Bypasser (rgthree), Image Comparer (rgthree), KSamplerSelect, LoadImage, mxSlider, Params In // RvTools, RandomNoise, Text Input [Dream], UNETLoader, UnetLoaderGGUF, UpscaleModelLoader, and VAELoader.
Workflow labels include workflow>Activate LoRA, workflow>Generate original, workflow>Get data, workflow>LoRA - 1, workflow>LoRA - 2, workflow>LoRA - 3, and workflow>Upscaler.
Toggle the built-in SeedVR2 upscaler on or off; each LoRA slot has individual on/off and strength controls.
For manual setup, the base option lists flux1-dev-fp8.safetensors or flux1-dev-fp16.safetensors with clip_l.safetensors; the GGUF option lists Q8, Q6, Q5, Q4, or Q3 choices for both the model and CLIP.
The auto version can download models automatically.
Before relying on the workflow, inspect its loaded connections in ComfyUI because the connections are not confirmed.
Suggestion
No license is specified in the listing; check UmeAiRT’s terms before use.
Suggestion
Does this need to be edited?
Sign in to send an edit request.
Sources
1 sourceSource excerpts
1 excerptSource context: 9770 downloads · Type Workflows · Base model Flux.1 D
Checkpoint · 11.8 GB · GGUF · fp16 · Hugging Face · UmeAiRT/ComfyUI-Auto_installer
Text encoder · 4.71 GB · Hugging Face · UmeAiRT/ComfyUI-Auto_installer
Text encoder · 308 MB · Hugging Face · UmeAiRT/ComfyUI-Auto_installer
BasicScheduler
PossibleNode pack · Unknown
DualCLIPLoaderGGUF
PossibleNode pack · Unknown
Fast Groups Bypasser (rgthree)
Not resolvedNode pack · Unknown
Image Comparer (rgthree)
Not resolvedNode pack · Unknown
KSamplerSelect
PossibleNode pack · Unknown
LoadImage
PossibleNode pack · Unknown
mxSlider
FoundNode pack · Unknown
Params In // RvTools
Not resolvedNode pack · Unknown
RandomNoise
Not resolvedNode pack · Unknown
Node pack · Unknown
UnetLoaderGGUF
PossibleNode pack · Unknown
UpscaleModelLoader
PossibleNode pack · Unknown
VAELoader
PossibleNode pack · Unknown
workflow>Activate LoRA
Not resolvedNode pack · Unknown
workflow>Generate original
Not resolvedNode pack · Unknown
workflow>Get data
Not resolvedNode pack · Unknown
workflow>LoRA - 1
Not resolvedNode pack · Unknown
workflow>LoRA - 2
Not resolvedNode pack · Unknown
workflow>LoRA - 3
Not resolvedNode pack · Unknown
workflow>Upscaler
Not resolvedNode pack · Unknown