Qwen I2I

Qwen Image Edit is Alibaba's image to image editing model, built on the 20 billion parameter Qwen Image backbone. Users provide an input image plus a natural language instruction, and the model produces an edited version that respects the prompt while preserving fidelity in untouched regions. It supports semantic edits such as changing object layout or orientation as well as appearance edits like style, texture, or detail changes. A standout feature is precise bilingual text editing in Chinese and English, where it can insert, replace, or tweak text inside the image while retaining font style, size, and layout. Architecturally, Qwen Image Edit uses a dual encoder approach that feeds the image into both a vision language head and a dedicated semantic encoder for stable visual edits.

August 19, 2025

Alibaba

Image to Image (I2I)