GitHub radar
Z-Image-Turbo — fast image synthesis in browser
Meet the new fast text-to-image generative model by Tongyi-MAI (Alibaba AI group), Z-Image-Turbo, which takes only 8 steps to generate images (compared to the standard 50 steps)! Also, a quick demo is here in the browser!
Z-Image-Turbo is an open text-to-image model developed by Tongyi-MAI(Alibaba AI group). It’s a compact version of Z-Image model with the same image quality but only requires 8 inference steps (compared to the standard 50). You could also test it out online at Hugging Face Space (without any GPU or registration): just input the description and specify the image size (512–2048 px) and the optional random seed to reproduce the result. This model supports English/Chinese text rendering in the image. And the weights are open source.
Why a vibe-coder should care
If you are looking for a good text-to-image generator but don’t want to pay for some service, then you could use this demo directly. Because it is really fast since it only requires 8 steps to generate images (compared to the standard 50)! If you like it, you could download the open weights and run it on your local machine with a powerful GPU.
How to install
Copy this and send it to your agent — Claude Code, Codex, any of them:
Open this demo: https://huggingface.co/spaces/mrfakename/Z-Image-Turbo — generate a test image from any description and show me the result
Nothing to install — runs in the browser.
Open on Hugging Face▌ More finds