Returning after a pause.
I left Civitai for a short time due to personal circumstances. However, during my absence, together with Triple Headed Monkey, we did not stop working and conducted several major projects.
Today I decided to publish one of the very first, but at the same time strong versions of our joint development — the improved ZIT base. Key changes compared to what was before: increased detail, reduced background blurring (less blurring) and improved realism.
Thanks for waiting. We are back and ready to continue working.
We use our training data and our soft for training.
!!! For best result use this TE THM_UNC_Z-image_TE - v1.0 | ZImage Text Encoder | Civitai
We released our another uncensored TE Z_Image_TE_SoLordZ - Z_Image_TE_SoLordZ_v1 | ZImage Text Encoder | Civitai
If you like our work and want support us please support us directly, since I am in a country which cannot get money from Civitai https://www.buymeacoffee.com/tripleheadg
our WF's SoLordZ WF (upscale) - v1.0 | ZImage Workflows | Civitai
and https://civarchive.com/models/2662358?modelVersionId=2989530
for realism use euler/normal, for art sometimes better result give euler_a/normal
All images in the gallery just raw generation w/o any tricks or upscale
ZIB SoLordZ v1
ZIB SoLordZ is a new chapter in the evolution of Z—Image and our collaboration with Triple Headed Monkey. The basic Z-Image model (up to 50 steps, CFG 5-9) is flexible, but requires fine tuning. Our goal was ambitious: to keep control of the "slow" model, but to teach it to generate in 8 steps with CFG=1.
We are releasing ZIB SoLordZ, one of our first public releases in a series of improvements to the Z Image Base.
ZIB SoLordZ is designed to work in "turbo mode": 8 steps, CFG=1. Due to this, it ignores the negative prompts. Everything that you want to exclude should be written directly in a positive prompt. This is especially important for anatomy, where control is crucial.
Description
Returning after a pause.
I left Civitai for a short time due to personal circumstances. However, during my absence, together with Triple Headed Monkey, we did not stop working and conducted several major projects.
Today I decided to publish one of the very first, but at the same time strong versions of our joint development — the improved ZIT base. Key changes compared to what was before: increased detail, reduced background blurring (less blurring) and improved realism.
Thanks for waiting. We are back and ready to continue working.
We use our training data and our soft for training.
!!! For best result use this TE THM_UNC_Z-image_TE - v1.0 | ZImage Text Encoder | Civitai
If you like our work and want support us please support us directly, since I am in a country which cannot get money from Civitai https://www.buymeacoffee.com/tripleheadg
for relism use euler/normal, for art sometimes better result give euler_a/normal
All images in the gallery just raw generation w/o any tricks or upscale
FAQ
Comments (24)
Excellent work! I love your content and I'm excited to see you back :D
Welcome back! 😉
hi, what's the difference between the different versions of 3-4b?
normal vs josiefied vs z-engineer-vs fedric?
Hi! "Normal" base TE model Qwen3-4b good also good but with version by fedric you get more details and his version TE better following the prompt
We've just uploaded our own Text Encoder model. If you'd like to compare that and give us your feedback on it, that would be greatly appreciated.
Can you please explain why all your examples have neg prompt for distilled zit with cfg 1?
Hi! :) This negative prompt does not affect anything because the model works on CFG=1, it remains in my WF from the Z Image Base. ComfyUI ignores that negative prompt anyway. So just ignore it :)
Review: 5-24-2026
This one does good in terms of realism. However, it doesn't do NSFW. Probably needs the xxx lora for that. The recommended fedric 3-4b is not as good as the josified. Day and night difference for my personal use. I'm interested to see where this one goes.
Thanks for the feedback. This is just the beginning. We have accumulated a lot of material and we have done a lot of work in various fields, mostly rebuilding the model from scratch. We will publish it gradually, and this is not the final stage yet. The final is just ahead
@speach1sdef178 thank you for the reply. i wasn't too harsh in the review as i read your situation and you probably missed a lot of stuff that has developed. It was only recently that the true "realism" made a breakthrough and that was with people mixing each other's checkpoints. Do you plan to develop an all-in-one or plan to do the sdxl way that has a strong base and allow loras to be used to allow the force use of the lora concept? So far in my experience with most of the checkpoints is that some do better than others in terms of prompt adherence. And some loras that allow a certain pose messes up the checkpoint. But so far zimage isn't really good at keyword concepts yet. it requires explaining in detail what one wants to see and even then its not guranteed.
@Melodic_Possible_582589 The Z Image is a specific model and it is quite difficult to manage for those who are used to working with previous generation models such as SDXL and requires some skill to create a prompt, but this model is worth paying attention to because it can do incredible things that other models simply are not capable of. We are working on all possible options, and I think the finale of our work will surprise many. I can't reveal any more information yet.
@Melodic_Possible_582589 Our plan is to build a strong base. We have trained a lot of topics, but we haven't gone into advanced poses etc. or anything truly XXX. We've left it soley at the anatomy on a professional level. And yes if you would like to do more with the model it will need custom lora on top.
@Triple_Headed_Monkey the problem of custom loras is they are trained on a vanilla ZIT and mess outputs of finetuned checkpoints.
@mphobbit Of course. If you want the best compatibility the lora should probably be trained on the same model.
@mphobbit I remember reading that if people trained the lora on z image base the problem would go away or lessen. But it still depends on the checkpoint as well. There are still a few checkpoints that get ruined when the lora is at a weight of 1.0.
Quick Review: May 29, 2026
I tested both recommended text encoders. This one focuses too much on warm/orange tones. It currently requires loras to do NSFW, otherwise it defaults to similar outputs as the original censored zimage turbo. Realism wise, it does well, but tends to have that grainy/fuzziness to it like the original zimage turbo.
It's neither a like or dislike for this checkpoint.
I want to see where this checkpoint goes.
I will post photos of what i got using the author's prompts. Portrait wise, it does well. But once you introduce another subject or poses things start to break.
@Melodic_Possible_582589 Hi, we uploaded another ours uncensored TE, you can try it to test if you want, I hope you'll like it :) Z_Image_TE_SoLordZ - Z_Image_TE_SoLordZ_v1 | ZImage Text Encoder | Civitai
using TE other than such the original censored one is useless ? (yes , no).
@amazingbeauty By combining our models and our TEs, you'll get the best results. The models also work with the base TE, but the results will be slightly different. We trained our TE specifically for our checkpoints, so the results are better when combined.
@speach1sdef178 sorry for commenting further , but is this added capabilities as fine-tune (the diffusion model) will affect how the zib originally abilities ? and a side question , what your thought on creating art posters or gaming content , go with zit or z-image base ?
@amazingbeauty No worries, and thanks for the question 🙂
The fine-tuning was designed to enhance ZIB's existing strengths rather than change or replace them. In fact, it improved many of the things ZIB was already good at. Styles have become cleaner, more distinct, and more recognizable, so the model's artistic identity is even stronger than before. The goal was to build upon ZIB's strengths, not to take them away.
As for posters and gaming-related content, I would personally recommend ZIB. This version is particularly strong when it comes to stylized artwork and creative illustrations. It also has good knowledge of many popular characters, making it very suitable for fan art, posters, and game-inspired content. In addition, ZIB works extremely well with LoRAs, which gives users even more flexibility and creative possibilities. I hope it helps you. I love art and game art and posters myself. We'll be releasing a couple more versions of ZIB soon.
Hi there,
I still wanted to leave my comment. I'm already familiar with your models and have no doubt about their quality, but I'd like to share a few nuances I personally encountered.
The model excels at creating realistic photos—it's clearly optimized for that, just like your previous models. However, I've personally run into an issue with styles. The prompts I usually use on your models tend to generate realistic images, and here too, people have mentioned that realistic outputs lean toward warm tones, which isn't always what I need.
So regarding styles: I personally struggle to get the results I'm aiming for. This isn't meant as criticism toward you—just an observation of what I've experienced. I don't know what to expect in future versions, or if there will even be any here, but I wanted to wish you luck and continued success in your work.
@Zakman99 Styles will be much more pronounced in later versions of ZIT. For style-focused work, ZIB is currently the better choice, as that model was originally designed around styles rather than realism.
ZIT, on the other hand, was originally built and tuned for realism. Although we have since changed and reworked virtually everything, most of those changes will appear in later versions.
The current version of ZIT is primarily focused on realism, with some support for styles. In future releases, ZIT will offer both strong realism and strong style capabilities, and the same will be true for future versions of ZIB — both realism and styles will be fully supported in both model lines.
For styles at this moment you can use our 1st ZIB model https://civitai.com/models/2655305/lord-zolaris?modelVersionId=2981612
Details
Available On (1 platform)
Same model published on other platforms. May have additional downloads or version variants.