This model is a precision-tuned ZiT merge designed to push the boundaries of fine-grained texture, anatomical accuracy, and atmospheric lighting while maintaining compatibility with LoRAs trained on the base models. I've train my character and style LoRAs on vanilla Z-Image Base. They work fine on Turbo derived models and tend to be more stable during training.
I will upload future models to Hugging Face also, it's a lot faster than dealing with a VPN
https://huggingface.co/PsychoLogic/PsychoMix_ZiT
The Composition
The architecture of this model is a blend of checkpoints and custom refinements:
v1.x
Core Foundation based on Moody Porn Mix v10 for subject pose, with various other models merged for style and consistency (see 'About this version' for specifics for a particular release)
For a more raw/realistic look I recommend using z-image-turbo-flow-dpo with a strength of -0.5 to -0.75 (yes, negative). It was baked into the base model and many (myself included) think it's merely a sloppifier.
v2.x
Core Foundation: A balanced merge of Z-Image Turbo NSFW v7.2 for subject pose, CyberRealistic Z-Image Turbo Catalyst v0.7 for style and technical precision, and selective merge from Famegrid ZIB v1.0, focused on later layers
Feedback & "Show Your Work"
I am actively looking for constructive feedback to improve this and other models. If the model isn't performing as expected for you, please help me fix it!
The "Pics or it didn't happen" Rule: If you encounter issues (artifacts, anatomy failures, or poor prompt adherence), please post your generation with metadata (Seed, Prompt, Sampler). Negative comments without an image and settings typically don't help. If you want a fix, show the failure so I can tune it out in the next merge.
Description
Trying something a bit more creative this time. This merge includes a large component of SeeSee21's Z-Anime, on top of Moody Porn Mix v10 and Juggernaut Z v1.0
FAQ
Comments (7)
looks good didn't work for me on neo forge ui, i used ae. and qwen_3_4b.safetensors that i sue with all z image turbo models but its just get me an error about mismatching
Which version? I'll spin up forge and have a look
I just tried v3.0 with forge and it worked fine. The v1.1 BF16 version is known to not work, but produces black outputs
@psychologicau which sampler and scheduler are you using on forge? I might have an older version but I don't have the res-2f_beta, i use euler beta and the likes and its over brigten.
@Melodic_Possible_582589 I primarily use ComfyUI, I find forge too restrictive. I've been testing with it using DPM++ 2M / SGM Uniform which seems OK. The base Moody model has DPO burned in, so you can undo some of its damage using negative weight for that LoRA if you don't want the slop filter. I tend to use -0.5 if I want something realistic. I will investigate adding this in at the merge level next time
I confirm, the checkpoint works for me too.
Neo has a bug: sometimes if you use for example FLUX checkpoint and then choose another model architecture from listbox (e.g. ZIT) the program keeps thinking you use FLUX checkpoint and tries to apply FLUX settings including vae and TE to ZIT checkpoint with obvious failed result. I had to reload the program to remove the issue.
Try to download a fresh version of Neo, the bug seems to be at least reduced.
@psychologicau i triedv1.2 it didn't work i didn't try v3.0 yet