Alibaba Qwen team's compact 7B-parameter open-weight model that unifies text-to-image generation and image editing in one checkpoint, with native RGBA transparency support, up to 10 reference images, and improved typography and portrait quality.
Qwen-Image-2.1 is an open-weight image model from the Qwen family that supports both text-to-image generation and image editing. Its visual generation component has 7 billion parameters, and it can produce RGBA images with transparency. Editing supports multiple reference images—up to 10—and localized instructions using circles, painted annotations, or masks. The model card documents local inference through the Diffusers pipeline, including examples for image generation, editing, transparent output, and 2K-resolution aspect ratios.
Qwen3-VL 8B is identified as the text encoder that encodes text instructions and condition images for Qwen-Image-2.1.
Diffusers provides the documented QwenImage21Pipeline used for local inference.
Be the first Cosign
Add it to your Products I use list. You can also share why you use it.
