Shoutout to China, they've given good competition and good local models!

From Qwen to now z-image,

Models like flux, and flux 2.0 being distilled and lobotomized and harder to finetune are not worth it anymore.


Currently been working on a finetune for Qwen on 200 Images which I'll release soon.

https://redd.it/1p7vnxo
@rStableDiffusion

The best thing about Z-Image isn't the image quality, its small size or N.S.F.W capability. It's that they will also release the non-distilled foundation model to the community.

## ✨ Z-Image
Z-Image is a powerful and highly efficient image generation model with 6B parameters. It is currently has three variants:

* 🚀 Z-Image-Turbo – A distilled version of Z-Image that matches or exceeds leading competitors with only 8 NFEs (Number of Function Evaluations). It offers ⚡️sub-second inference latency⚡️ on enterprise-grade H800 GPUs and fits comfortably within 16G VRAM consumer devices. It excels in photorealistic image generation, bilingual text rendering (English & Chinese), and robust instruction adherence.

* **🧱 Z-Image-Base – The non-distilled foundation model. By releasing this checkpoint, we aim to unlock the full potential for community-driven fine-tuning and custom development.**

* ✍️ Z-Image-Edit – A variant fine-tuned on Z-Image specifically for image editing tasks. It supports creative image-to-image generation with impressive instruction-following capabilities, allowing for precise edits based on natural language prompts.

**Source:** https://www.modelscope.cn/models/Tongyi-MAI/Z-Image-Turbo/

https://redd.it/1p7ykw8
@rStableDiffusion