Z-Image: 一种基于单流扩散Transformer的高效图像生成基础模型
Z-Image: An Efficient Image Generation Foundation Model with Single-Stream Diffusion Transformer
AI总结：
提出Z-Image，一种6B参数的高效图像生成基础模型，采用可扩展单流扩散Transformer架构，通过优化数据基础设施和训练流程，仅用314K H800 GPU小时完成训练，性能媲美顶级商业模型。
摘要：
高性能图像生成模型领域目前由专有系统主导，例如Nano Banana Pro和Seedream 4.0。领先的开源替代方案，包括Qwen-Image、Hunyuan-Image-3.0和FLUX.2，其特点是参数数量庞大（20B到80B），使得它们在消费级硬件上进行推理和微调变得不切实际。为了解决这一差距，我们提出了Z-Image，一个高效的6B参数基础生成模型，基于可扩展单流扩散Transformer（S3-DiT）架构，挑战了“不惜一切代价扩大规模”的范式。通过系统地优化整个模型生命周期——从精心策划的数据基础设施到精简的训练课程——我们仅用314K H800 GPU小时（约63万美元）就完成了完整的训练流程。我们的少步蒸馏方案结合奖励后训练进一步产生了Z-Image-Turbo，在企业级H800 GPU上提供亚秒级推理延迟，并与消费级硬件（<16GB VRAM）兼容。此外，我们的全预训练范式还使得能够高效训练Z-Image-Edit，一个具有令人印象深刻的指令跟随能力的编辑模型。定性和定量实验均表明，我们的模型在多个维度上达到了与领先竞争对手相当或超越的性能。最值得注意的是，Z-Image在逼真图像生成和双语文本渲染方面表现出卓越的能力，其结果可与顶级商业模型相媲美，从而证明了以显著降低的计算开销实现最先进结果是可行的。我们公开发布了我们的代码、权重和在线演示，以促进可访问、预算友好且最先进的生成模型的发展。
英文摘要：
The landscape of high-performance image generation models is currently dominated by proprietary systems, such as Nano Banana Pro and Seedream 4.0. Leading open-source alternatives, including Qwen-Image, Hunyuan-Image-3.0 and FLUX.2, are characterized by massive parameter counts (20B to 80B), making them impractical for inference, and fine-tuning on consumer-grade hardware. To address this gap, we propose Z-Image, an efficient 6B-parameter foundation generative model built upon a Scalable Single-Stream Diffusion Transformer (S3-DiT) architecture that challenges the "scale-at-all-costs" paradigm. By systematically optimizing the entire model lifecycle -- from a curated data infrastructure to a streamlined training curriculum -- we complete the full training workflow in just 314K H800 GPU hours (approx. $630K). Our few-step distillation scheme with reward post-training further yields Z-Image-Turbo, offering both sub-second inference latency on an enterprise-grade H800 GPU and compatibility with consumer-grade hardware (<16GB VRAM). Additionally, our omni-pre-training paradigm also enables efficient training of Z-Image-Edit, an editing model with impressive instruction-following capabilities. Both qualitative and quantitative experiments demonstrate that our model achieves performance comparable to or surpassing that of leading competitors across various dimensions. Most notably, Z-Image exhibits exceptional capabilities in photorealistic image generation and bilingual text rendering, delivering results that rival top-tier commercial models, thereby demonstrating that state-of-the-art results are achievable with significantly reduced computational overhead. We publicly release our code, weights, and online demo to foster the development of accessible, budget-friendly, yet state-of-the-art generative models.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
