Qwen Image 2.1 released

Alibaba's Qwen team open-sourced Qwen-Image-2.1, a compact 7B-parameter model that unifies text-to-image generation and image editing in one checkpoint. It adds native RGBA transparency support, accepts up to 10 reference images for identity-preserving edits, and supports mask- and annotation-based local editing, all usable directly through the Diffusers library. Qwen also shipped two companion prompt-rewriting models alongside it: one that turns vague image-editing instructions plus reference images into precise, structured edit prompts, and another that expands short, any-language text-to-image requests into detailed English prompts with a recommended aspect ratio. Together the release gives teams building self-hosted or local image pipelines a compact, open-weight alternative that handles both generation and editing without stitching together separate models, and the prompt-rewriting layer addresses a common failure point where short or vague prompts produce weak results.