Loading
Vous avez un workflow ComfyUI ? Obtenez une documentation lisible directement dans votre navigateur.Rendre un workflow lisible.Essayer maintenant ↗
README: XUPLX ZIMAGE MERGE + INPAINT + IMG2IMG + UPSCALE WORKFLOW
Profil d'exécution
Description de la source
README: XUPLX ZIMAGE MERGE + INPAINT + IMG2IMG + UPSCALE WORKFLOW
Base UNet checkpoint: z_image_turbo_bf16.safetensors
CLIP: qwen_3_4b.safetensors with lumina2
VAE: ae.safetensors
Main latent canvas is 1024×1536 (SD3 latent node), so this workflow is set up to generate portrait-format images at that size by default.
Commentaire généré par l’IA
Explication générée par l’IA à partir des détails de la source et de la configuration. Les suggestions sont clairement signalées.
Use XUPLX_INPAINT_IMG2IMG-REFINEMENT_UPSCALE to create portrait images from MOOD and NATURAL LANGUAGE prompts, then inpaint a merged image, refine it with img2img, and upscale it to 2048×3072. Inputs include the prompts, a merged image, and an inpaint prompt; the output is an up4
The MOOD prompt runs early and NATURAL LANGUAGE later; ModelSamplingAuraFlow adjusts their blend in the continuing latent branch.
After generation, load the merged image for inpainting. Img2img refines it, and SeedVR2 upscales the img2img output.
Enter text in the MOOD and NATURAL LANGUAGE prompt inputs. For inpainting, load the merged image and add a prompt describing the change.
The generation canvas defaults to portrait 1024×1536; SeedVR2 is configured for 2048×3072 output in lab color mode.
Required node packs include ComfyUI_Comfyroll_CustomNodes, comfyui_layerstyle, comfyui-crystools, comfyui-easy-use, and comfyui-rvtools_v2.
Also required are rgthree-comfy, comfyui-mixlab-nodes, pr-was-node-suite-comfyui-47064894, and seedvr2_videoupscaler.
Named model files are ae.safetensors, ema_vae_fp16.safetensors, qwen_3_4b.safetensors, seedvr2_ema_7b-Q5_K_M.gguf, and z_image_turbo_bf16.safetensors.
The defaults use shared seed 01234567890 and 20 global steps: stage one runs from 0–5 and stage two from 5–20.
ModelSamplingAuraFlow is set to 5. The optional XUPLX_UglyPeopleLoRA.safetensors is on by default at strength 0.4 / 0.4 in the MOOD stage; the inpaint branch uses the base model path.
Img2img denoise defaults to 0.45 and may vary between images.
For upscaling, the workflow suggests GGUF models. If VRAM purge fails, an out-of-memory error may occur; try the upscale again.
Suggestion · non vérifié
Faut-il modifier cela ?
Connectez-vous pour envoyer une demande de modification.
Sources
1 sourceExtraits de sources
1 extraitSource context: 978 downloads · Type Workflows · Base model ZImageTurbo
Workflow generates MOOD early, then NATURAL LANGUAGE late, and you can tune the “blend” between them.
Mechanically, that’s implemented as:
A single fixed seed shared across the samplers (seed = 01234567890).
A single global STEPS integer (set to 20) feeding multiple samplers.
A shared END_AT/START_AT integer (set to 5), so stage 1 runs 0→5 steps and stage 2 runs 5→20 steps.
The “blend knob” is ModelSamplingAuraFlow set to 5 in the branch that continues the latent (the README refers to this as “somewhere-in-the-middle” tuning).
MOOD and NATURAL LANGUAGE prompts are centralized as Mixlab TextInput_ nodes (“MOOD” and “NATURAL LANGUAGE”), which drive the CLIPTextEncode nodes via links.
Load an optional LoRA (XUPLX_UglyPeopleLoRA.safetensors); strength is set to 0.4 / 0.4, switch On by default.
The LoRA-modified model is used in the “MOOD” portion of the workflow, while other parts (notably the inpaint branch) are using the base model path. The XUPLX LoRA is optional and trained on real photographs.
Download XUPLX_UglyPeopleLora Here:
After generation, you do a classic inpaint pass:
Load your merged image (simply copy and paste into "Load Image" node).
Inpaint prompt example in the workflow: “The woman is wearing a black colored backpack.”
Img2Img is meant to synthesize the lighting, ambiance, and subject from the merged image with the inpainted material. Denoise is set to 0.45 by default, but varies from one image to another.
It’s configured to output 2048×3072 and "lab" color mode by default.
The VAE for SeedVR2 is set to ema_vae_fp16.safetensors. It is strongly suggested to use GGUF models for upscaling. You may see an OOM error if VRAM purge fails. Simply run the upscale again.
This workflow expects several custom packs/plugins to be installed, including
Mixlab nodes (TextInput_)
Comfyroll (CR Load LoRA)
RvTools v2 (Integer nodes for STEPS and END_AT/START_AT)
rgthree (Image Comparer)
WAS Node Suite (Seed node)
SeedVR2 upscaler pack
The workflow comes with a README , or you can check out the article I wrote :
How to Merge Image Styles Quickly with XUPLX and ZImage Turbo
You can download a simplified version of my workflow here:
IF YOU LOVED THIS...DONT BUY ME A COFFEE!!
Download " ExportGPT - Media & Chat Export Toolkit " in the Chrome Web Store. If you live in ChatGPT and Sora, it's an indispensable tool to move your assets around quickly without external servers. IT HAS A FIVE STAR RATING AND IT'S FREE FOREVER!
See it in action: exportgpt.org
Remaining hardware or download-size estimates are not available yet.
Exigences
ae.safetensors
TrouvéVAE · 320 MB · SAFETENSORS · Unknown
VAE · 478 MB · SAFETENSORS · fp16 · Unknown
Text encoder · 7.49 GB · SAFETENSORS · Unknown
seedvr2_ema_7b-Q5_K_M.gguf
Non résoluCheckpoint · Unknown
seedvr2_videoupscaler
Non résoluUpscaler · Registry
Checkpoint · 11.5 GB · SAFETENSORS · bf16 · Unknown
ComfyUI_Comfyroll_CustomNodes
TrouvéPack de nœud · Registry
comfyui_layerstyle
TrouvéPack de nœud · Registry
comfyui-crystools
TrouvéPack de nœud · Registry
comfyui-easy-use
TrouvéPack de nœud · Registry
comfyui-mixlab-nodes
PossiblePack de nœud · Registry
Pack de nœud · Registry
pr-was-node-suite-comfyui-47064894
Non résoluPack de nœud · Registry
rgthree-comfy
TrouvéPack de nœud · Registry
