you ‧ notes ‧ unread ‧ starred ‧ add url

Qwen Image 2.1

september 2026 ‧ ai imagegen models ‧ open in karakeep

by Qwen Team ‧ Qwen ‧ published 2026.09.20 ‧ saved copy from 2026.09.22

Qwen-Image-2.1 is an open-source image model featuring a compact 7B parameter architecture that unifies text-to-image generation and image editing, including native support for transparent images. It offers efficient inference, multi-reference editing (up to 10 images), flexible local editing with masks or annotations, and improved fidelity in preserving people’s identity and product features. The model supports diverse tasks such as panorama, infographic, and storyboard creation while enhancing visual quality with realistic textures and refined typography. Overall, Qwen-Image-2.1 is designed for versatile and cost-effective image generation and editing in various creative and commercial applications.

regenerate summary

page description:

GITHUB HUGGING FACE MODELSCOPE We are excited to open-source Qwen-Image-2.1, an image model in the Qwen family that balances generation quality, inference efficiency, and cost. Qwen-Image-2.1 unifies text-to-image generation and image editing in a single model, with just 7B parameters in its visual generation component and native support for generating and editing transparent images. This update introduces four key improvements: Compact and efficient: A lightweight architecture and inference optimizations balance image quality and computational cost.

« back to bookmarks