Chroma Flux | ComfyUI Tweaked | Workflow
Basic ComfyUI workflow for Chroma Diffusion
Profil d'exécution
- Accès
- Flux de travail externe
- Statut
- Présenté
- Source
- Civitai
- Mis à jour
- 25 oct. 2026
Description de la source
Basic ComfyUI workflow for Chroma Diffusion
Necessary extension , put into "ComfyUI\custom_nodes" folder:
https://github.com/lodestone-rock/ComfyUI_FluxMod
Models folder same as for Flux:
"ComfyUI\models\unet"
"ComfyUI\models\clip"
Chroma GGUF models run same as normal ones, use "Load Chroma Diffusion Model" and "Padding Removal" custom nodes with T5 text encoder only.
Checkpoints for GPU's with high VRAM :
https://huggingface.co/lodestones/Chroma/tree/main
Checkpoints for GPU's with medium VRAM :
https://huggingface.co/silveroxides/Chroma-GGUF/tree/main
T5 text encoder (e4m3fn preferred for medium VRAM):
https://civitai.com/models/497255?modelVersionId=568405
Suggested parameters to use:
I prefer euler ancestral with beta scheduler as it follows prompt better than just euler beta or other combinations as I experimented.
It produces a lot of dots look like trypophobia hell and noise clusters at image edges. So "ModelSamplingFlux" with "max_shift" = 1.5 "base_shift" = 0.5 improves quality pretty much.
Other parameters same as for Flux:
good results at 25 steps with CFG 4.0-5.0, faster at 20 steps
What you need
Hardware and download-size estimates are not available yet.
Exigences
Modèles
5chroma-unlocked-v16-Q4_0.gguf
Non résoluCheckpoint · Unknown
CLIPLoader
Non résoluText encoder · Unknown
stableDiffusion3SD3_textEncoderT5E4m3fn.safetensors
Non résoluText encoder · Unknown
VAEDecode
PossibleVAE · Unknown
VAELoader
Non résoluVAE · Unknown
Packs de nœuds
7ChromaDiffusionLoader
Non résoluPack de nœud · Unknown
ChromaPaddingRemoval
Non résoluPack de nœud · Unknown
CLIPTextEncode
PossiblePack de nœud · Unknown
EmptyLatentImage
PossiblePack de nœud · Unknown
KSampler
PossiblePack de nœud · Unknown
- Trouvé
Pack de nœud · Unknown
SaveImage
PossiblePack de nœud · Unknown
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 generating Chroma Diffusion images from a text prompt, using a Chroma checkpoint and T5 text encoder; its named VAEDecode and SaveImage components suggest decoded images can be saved as output.
Suggestion · non vérifié
Entrées et sorties
For Chroma GGUF models, the description specifies a T5 text encoder only and the custom nodes “Load Chroma Diffusion Model” and “Padding Removal.”
Choose a Chroma checkpoint from the listed high- or medium-VRAM collections; e4m3fn is preferred for the T5 encoder on medium VRAM.
Installation
Install the ComfyUI_FluxMod extension in ComfyUI\custom_nodes.
Place model files in ComfyUI\models\unet and ComfyUI\models\clip.
Dépendances
Named model files include chroma-unlocked-v16-Q4_0.gguf and stableDiffusion3SD3_textEncoderT5E4m3fn.safetensors.
Other named components are ChromaDiffusionLoader, ChromaPaddingRemoval, CLIPLoader, CLIPTextEncode, EmptyLatentImage, and KSampler.
The named list also includes ModelSamplingFlux, SaveImage, VAEDecode, and VAELoader.
Configuration
Try Euler ancestral with a beta scheduler as a starting setting; the description’s author prefers it to Euler beta and other combinations.
Suggestion · non vérifié
If dot-like marks or edge noise appear, try ModelSamplingFlux with max_shift = 1.5 and base_shift = 0.5.
Suggestion · non vérifié
The description suggests 25 steps with CFG 4.0–5.0 and also mentions 20 steps as an option.
Suggestion · non vérifié
Précautions
No model-use license is listed; check the applicable usage terms before using the checkpoint or encoder.
Suggestion
Faut-il modifier cela ?
Connectez-vous pour envoyer une demande de modification.
Sources
1 source
Sources
1 sourceExtraits de sources
1 extrait
Extraits de sources
1 extraitSource context: 453 downloads · Type Workflows · Base model Flux.1 S
