Aller au contenu
WorkflowPrésentéExterne
Ouvrir la source

Starnodes Example Workflow Flux + Upscaler + Refiner HQ — v1.0

This is my Example Workflow using the new Starnodes. Just open the workflow and install the missing nodes in the comfyui manager or visit the github:

Sourcecivitai.com/models/1018596
TâcheMise à l'échelleFamilleFLUXVersionRelease 1142105 — Example_Workflow_Flux_Refiner_With_Starnodes.jsonEstimé VRAMNot yet characterizedBClasse

Profil d'exécution

Accès
Flux de travail externe
Statut
Présenté
Version
Example_Workflow_Flux_Refiner_With_Starnodes.json
Source
Civitai
Mis à jour
18 oct. 2026

Description de la source

This is my Example Workflow using the new Starnodes. Just open the workflow and install the missing nodes in the comfyui manager or visit the github: https://github.com/Starnodes2024/ComfyUI_StarNodes

This workflow uses upscaling and refining.

The nodes include:

NEW: ⭐ Detail Star Deamon is adapted from the original sources. This can improve the details of your generated images and can be connected to the Starsamplers. Works with Flux and all SD Models. Read more about the settings and how it works on the original sources: https://github.com/muerrilla/sd-webui-detail-daemon / https://github.com/Jonseed/ComfyUI-Detail-Daemon

NEW: ⭐ StarSampler SD / SDXL: A Ksampler for SD, SDXL, SD3.5 with settings and outputs for model and conditioning passtrough. Optional: Connector for Detail Star Deamon

NEW: ⭐ StarSampler FLUX: A Ksampler for Flux with settings and outputs for model and conditioning passtrough. Optional: Connector for Detail Star Deamon

⭐ SD(XL) Starter : Is loading checkpoint (with CLIP and VAE) and create an empty latend (you can choose resolutions or set your own)

⭐ FLUX Starter : Is loading Unet (Diffusion Model), 2 Clips and create an empty latend (you can choose resolutions or set your own)

⭐ SD3.0/3.5 Starter: Is loading Unet (Diffusion Model), 3 Clips and create an empty latend (you can choose resolutions or set your own)

*you can add more ratios if you edit the .json-files in the node folder.

⭐ Star Model Latent Upscaler: Decode an input Latent and Vae to an Image, upscale with choosen model, resize to given size, decode output image back to laten with selected VAE.

⭐ Star Seven Wildcards: A prompt maker that use 7 inputs with wildcards to create prompts with different wildcards and multiple random inputs from one wildcard used. Also works without wildcards as text concatenate. You will find many wildcards in the "wildcards" subfolder of the node. To use them just copy the whole folder to your main ComfyUI directory.

⭐ Input Image Chooser: A simple switch for big workflows to switch between 8 input images

What you need

Model files
2.9 GiB known subtotal

Remaining hardware or download-size estimates are not available yet.

Exigences

Modèles

6

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.

Objectif

  • This ComfyUI workflow is for creating and refining images with FLUX, SD, SDXL, or SD3.0/3.5: choose a model setup and resolution, provide text or wildcard prompt inputs, and work with image and latent data for detail refinement or upscaling.

Flux

  • Star Seven Wildcards combines seven wildcard or plain-text inputs into prompts, including multiple random entries from one wildcard.

  • Input Image Chooser switches among eight input images.

  • Detail Star Deamon can connect to the StarSamplers for detail refinement and is described as working with Flux and all SD models.

Entrées et sorties

  • The FLUX starter loads a diffusion model and two CLIP encoders; the SD3.0/3.5 starter loads a diffusion model and three CLIP encoders; and the SD(XL) starter loads a checkpoint with CLIP and VAE. Each creates an empty latent.

  • Star Model Latent Upscaler takes a latent, VAE, and upscaling model, converts the latent to an image, upscales and resizes it, then converts it back to latent data with a selected VAE.

  • StarSampler FLUX and StarSampler SD / SDXL pass through model and conditioning outputs; the SD / SDXL sampler supports SD, SDXL, and SD3.5.

Installation

  • If nodes are missing, install them through ComfyUI Manager or visit the ComfyUI_StarNodes GitHub repository.

    Suggestion · non vérifié

  • To use the wildcard files, copy the full wildcards subfolder into your main ComfyUI directory.

    Suggestion · non vérifié

Dépendances

  • Node names listed with the workflow include DetailStarDaemon, Fluxstarsampler, FluxStartSettings, LoraLoaderModelOnly, PreviewImage, SaveImage, ShowText|pysssss, StarFiveWildcards, Starupscale, and VAEDecode.

  • Model files named for the workflow include 4x_NMKD-Siax_200k.pth, ae.safetensors, flux1-canny-dev-lora.safetensors, FluxDevTurbo8Steps.safetensors, t5xxl_fp16.safetensors, and ViT-L-14-TEXT-detail-improved-hiT-GmP-HF.safetensors.

Configuration

  • To get started, choose a FLUX, SD(XL), or SD3.0/3.5 starter, then use a listed resolution or set a custom one.

    Suggestion · non vérifié

  • You can add resolution ratios by editing the .json files in the node folder, or connect Detail Star Deamon to the StarSamplers.

    Suggestion · non vérifié

Précautions

  • The listed files 4x_NMKD-Siax_200k.pth, FluxDevTurbo8Steps.safetensors, and t5xxl_fp16.safetensors have no download URL in the listing; check where to obtain them before use.

    Suggestion · non vérifié

Faut-il modifier cela ?

Connectez-vous pour envoyer une demande de modification.

Sources

1 source
CivitaiPrincipalVersion 1142105
civitai.com/models/1018596

Vérifié 18 oct.

Extraits de sources

3 extraits
Note18 oct.civitai.com/models/1018596

Source context: 403 downloads · Type Workflows · Base model Flux.1 D

Note4 aoûtcivitai.com/models/1018596

Source context: 367 downloads · Type Workflows · Base model Flux.1 D

Note20 juil.civitai.com/models/1018596

Source context: 350 downloads · Type Workflows · Base model Flux.1 D