Qwen-Image is an open-source image foundation model by Alibaba’s Qwen team, with ~20 billion parameters, built on the MMDiT (Multimodal Diffusion Transformer) architecture. It makes significant advances in complex text rendering (including mixed Chinese/English, multi-line/multi-paragraph text) and precise image editing, supporting both generation and editing of diverse visual content.
Please login to use description