Home
Creators
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Register Favorites Keys Review DMs Logout
Community
Telegram ThePornDude
Home Creators Posts Import Register Favorites Logout

Z-IMAGE BASE ULTRA NEW TEXT TO IMAGE AI KING WORKFLOW V2! (Patreon)

Published:
2026-01-31 20:31:36
Edited:
2026-02-07 19:48:02
Imported:
2026-06
Tags:

Downloads

Content

UPDATE 7 february 2026: Hey everyone! I just released a new video update for the V2 Workflow that you can watch here: https://youtu.be/PAh870We7tIIn this update I added a ControlNet workflow for better pose and composition control, a Detailer Daemon pass to crank the micro details even higher, AND THE 8 STEPS LORA that allows you to generate images in 8 steps instead of 25!

Just use the Z-IMAGE_BASE_ULTRA-MODELS-NODES_INSTALL-V2.bat installer from this post:https://www.patreon.com/posts/z-image-base-new-149607634put it in your comfyui folder, run it to install the missing nodes and files, then update comfyui to the latest version with the update comfyui bat file inside the update folder then download the the updated workflow from this post drop it into your ComfyUI folder and you’re ready to go 😉

Hey everyone! I’ve just released a new video on THE NEW TEXT TO IMAGE AI KING BASE MODEL called “Z-IMAGE”. The BIG difference compared to Z-IMAGE-TURBO is that this is the BASE model, that has amazing  image image diversity and styles, meaning that real finetuning and training is now possible, so the community can finally cook up new LoRAs, styles, characters and insane custom versions while running with LESS THAN 8GB OF VRAM! JUST INSANE! 

You can check out the video right here: https://youtu.be/9EeCWOJvLOw

I also created a 1-click installer for the Webui for my Patreon supporters for Local AND Runpod, that you can find right here: https://www.patreon.com/posts/149607634

Here the link to the manual install guide: https://pastebin.com/6HBikbsX

Down below you will also find the Z-IMAGE_BASE_ULTRA_WORKFLOW-V2.json for you to use!

As always, supporting me on Patreon allows me to keep creating helpful resources like this for the community. Thank you for your support - now go have some fun😉!

Files

Comments

Virtamouse

Why are there 2 lora nodes for the I2I section? I get really grainy looking images for the ultra combo section.

skyneta

same happening for me too

Kaldo

Top, comme d'habitude. Merci !

Albion22

try reducing the grain intensity at the Z-image combo TXT Grain box

Virtamouse

Ive turned it off. Same result. The non combo i2i looks much better.

Tobias Gampe

I appreciate your work. It would be even better if the prompts and images were available online, to better understand the workflow.

Ryan Prosser

Overall this is really amazing! For one part, Don't know if it's just me, but the Z-image (Image to Image) takes into account the image uploaded, but when I use "Z-IMAGE ULTRA COMBO IMAGE TO IMAGE" it appears to completely ignore the uploaded image

Virtamouse

lower your denoise to around 50-60

Shap

Great job, thanks! One question about the ultra combo workflow: I would like the first pass of base model to be shorter and the second pass of turbo model to be longer -> do I have to change something in the subgraph node settings to do that, how? Thanks in advance ;)

Super Assetto Corsa

Z-Image Turbo still better at following the prompt, I think I'll keep using Turbo, it's also faster and almost same quality as Base model.

Edward Rocha

Any help I'm trying to run the it but I'm getting UnetLoaderGGUF --> GGUF magic invalid, I used the AUTO_INSTALL.bat

Oscar Petoskey

This is strange, Z-Image Combo works perfectly in both T2I and I2I. BUT, Z-Image base, pukes with error: Prompt execution failed Prompt outputs failed validation: VAEEncode: - Required input is missing: vae VAEDecode: - Required input is missing: vae

Martin Gee

Great Work! <3 Is there any chance you are adding IPAdapter and/or ControlNet to the image to image workflow?

Shawn Ruegge

I have a couple of loras trained on base but the faces aren't even close when I use the combo T2I.

Elsa

Hey! Love the Workflow 💗 But why did you change the the sampler from res_multistep to euler in your comby workflows? specially when you said in your video that you switched to sa_solver_pece except performance wise back to res_multistep? is euler needed for the detail-deamon to work or whats the reason?

Dmoe2600

I'm getting missing nodes FastFilmGrain and FastLaplacianSharpen

Michael

Something is wrong with the new workflow and the Lora: lora key not loaded: lora_unet__context_refiner_0_attention_to_k.alpha lora key not loaded: lora_unet__context_refiner_0_attention_to_k.lora_down.weight lora key not loaded: lora_unet__context_refiner_0_attention_to_k.lora_up.weight Basically it's that long error that says hey this lora does not match the model. Also what Elsa said below - on some of your other sections you changed settings like 0.7 vs 0.6 weight on the Lora Power Loader as well as changing the sampler. Anyway, just wanted to report some issues. It's still generating but I feel like this can be tweaked to work even better.

Eric Balfour

why all image are anime, cartoon rendered ? and very contrasting. and how work the inpainting, where i can load an alpha or paint ? thanks

Poppa Joe

Having the same problem, and cant find how to install them

Tribaloide

I get this error: Loop (69,577,70) - not submitting workflow

Michael Young

getting the same with the V2 workflow, you find any solution?

Tribaloide

No. I haven't found a solution; the same error keeps appearing. I hope its creator can help us resolve it.

The Monster Girls Team

I dragged the workflow v2 into comfyui, it found and hooked everything up, and when I go to generate, I just get grainy / weird renders.. can we have samples when you release things? seed, text prompt, and settings (dimensions, etc) so we can try to reproduce exactly the same thing, to be sure we didn't make a mistake?

Matt

King

Matros

Why do I get black output for everything? It just doesn't work for me! I have RTX5090.

Matros

The problem is that all processes complete without a single error, yet the final output is completely black with no image. I couldn't find any error in the log file or in the workflow, even with AI assistance. I have no idea what could be causing it.