Alibaba Unveils Qwen-Image Model for Advanced Text Rendering and Image Editing

Alibaba Group has released Qwen-Image, a new image foundation model featuring 20 billion parameters. This model is designed to achieve significant advancements in complex text rendering and precise image editing. Users can access Qwen-Image through Qwen Chat by selecting the 'Image Generation' option.

Qwen-Image excels in rendering intricate texts with high precision, supporting both alphabetic languages like English and logographic languages such as Chinese. It handles multi-line layouts, paragraph-level semantics, and fine-grained details with high fidelity. The model also offers consistent image editing capabilities, preserving semantic meaning and visual realism through an enhanced multi-task training paradigm.

The model has been evaluated on multiple public benchmarks, consistently outperforming existing models in both image generation and editing tasks. It is available on open-source platforms such as Hugging Face, GitHub, and Alibaba's ModelScope, providing developers with a robust tool for creating AI-driven applications. Alibaba invites community participation and feedback to further develop an open and sustainable generative AI ecosystem.