Z Image

Z-Image is a powerful AI model specifically designed with strong capabilities in photorealistic image generation, accurate rendering of both Chinese and English text, and robust adherence to bilingual instructions. It achieves exceptional performance comparable to or exceeding leading competitors with only 8 steps. The model adopts an innovative Scalable Single-Stream DiT (S3-DiT) architecture, unifying various conditional inputs into a single sequence for maximum parameter efficiency. It performs exceptionally well in image generation, with most generations taking a maximum of 2 seconds on NVIDIA A10 GPUs and 2-5 seconds on high-end to mid-range consumer GPUs.