Tongyi Qianwen unveiled on August 5 the open-source release of Qwen-Image, marking the debut of its series of image generation foundation models. Built on an MMDiT architecture with 20 billion parameters, Qwen-Image has demonstrated substantial advancements in rendering complex texts and editing images with precision. The model excels in text rendering, consistent image editing, and robust cross-benchmark performance. Supporting multi-line layouts, paragraph-level text generation, and fine-grained detail presentation, Qwen-Image delivers high-fidelity outputs for both English and Chinese, achieving state-of-the-art (SOTA) results in numerous public benchmark tests.