Z Image Turbo

Z Image Turbo is listed under an enterprise pricing model, so expect to talk to the team rather than pick a self-serve plan off a pricing page.

It is an AI image generator that turns text prompts into photorealistic pictures, and speed is the main selling point.

The model has 6B parameters and uses a Decoupled-DMD technique that needs only 8 inference steps, which is how it reaches sub-second generation times.

Prompts can be written in English or Chinese, which matters if you work across both languages and want text in images to come out right.

Z Image Turbo is also built to run on consumer GPUs with 16GB of VRAM, so you do not need a server rack to try it locally.

As a text to image tool it is geared toward realistic output more than stylized art, and that focus is its clearest strength.