Z Image Turbo – Ultra-fast 2K AI image generator with bilingual text

I built a small web app for trying out Z Image Turbo, an ultra-fast text-to-image model that generates up to 2K (2048×2048) images and handles English + Chinese text in a single shot: https://zimageturbo.app/

A few details: – Uses the 6B-parameter Z Image Turbo model from Tongyi-MAI, which fits in 16GB VRAM while still producing high-quality images. – Supports up to 2048×2048 output and multiple formats (PNG/JPG/WebP). – Optimized for prompts that mix image description and typography (e.g. posters, covers, logos with both English and Chinese text).

My goal with this site is: – Make it easy to try the model without setting up GPUs, ComfyUI, or downloading weights. – Provide a simple, focused UI for fast iteration instead of a “do everything” image lab.

I’d love feedback from HN on: – Latency and UX (does it feel “instant” enough for you?). – Prompt handling, especially for bilingual text. – What knobs / features you’d actually want exposed (seed control, negative prompts, presets, etc.).

Link: https://zimageturbo.app/

Happy to answer technical questions in the comments and share more details about how it’s wired up.