---
license: creativeml-openrail-m
base_model: Tongyi-MAI/Z-Image-Turbo
library_name: mlx
tags:
- mlx
- mflux
- text-to-image
- z-image
- civitai
---

# CyberRealistic Z-Image Turbo v4.0 — MLX (8bit)

> Part of the [CyberRealistic Z-Image-Turbo v4 · mflux](https://huggingface.co/collections/avlp12/cyberrealistic-z-image-turbo-v4-mflux-6a565b41c2e1c0c5701b0436) collection.

[CyberRealistic Z-Image Turbo](https://civitai.com/models/2218365) by
[Cyberdelia](https://civitai.com/user/Cyberdelia) (v4.0, from the creator's own
[HF mirror](https://huggingface.co/cyberdelia/cyberrealistic_zimage)), converted from the ComfyUI
single-file checkpoint to the [mflux](https://github.com/filipstrand/mflux) layout and quantized to
**8bit** for Apple-silicon Macs. Text encoder (Qwen3-4B), VAE, and tokenizer come from the base
[Tongyi-MAI/Z-Image-Turbo](https://huggingface.co/Tongyi-MAI/Z-Image-Turbo) (Apache-2.0).

A photorealism-focused finetune of Z-Image-Turbo: ~9 steps, no CFG, multilingual prompts
(Korean included). The 4-bit build runs on a 16 GB Mac.

```python
from mflux.models.z_image.variants import ZImageTurbo
m = ZImageTurbo(model_path="avlp12/CyberRealistic-Z-Image-Turbo-v4-mflux-8bit")
m.generate_image(seed=7, prompt="candid photo of ...", num_inference_steps=9).save(path="out.png")
```

Used as a built-in model in [Alis Studio](https://github.com/avlp12/alis-studio).
License: CreativeML OpenRAIL-M (inherited from the finetune; see the use restrictions).
Credit: Cyberdelia (finetune), Tongyi-MAI (base model), mflux (MLX runtime).