Alibaba: qwen-image-edit-plus API

Qwen-Image-Edit-Plus is a high-performance image editing model from Alibaba's Tongyi Lab, built on a 20B-parameter multimodal diffusion architecture. A significant upgrade over the standard version, its core strengths lie in "multi-image fusion" and "ultra-high-precision text preservation." It supports simultaneous input of up to 3 images (e.g., a face, an outfit, and a background) for seamless visual synthesis based on text instructions. With integrated ControlNet technology, it provides precise control over object rotation, pose transformation, and lighting reconstruction. Particularly in advertising and e-commerce design, it enables the modification of image content while fully preserving the original Chinese/English fonts, styles, and layouts.

• Context window: 2,048 tokens
• Input: text, image
• Output: image
• File input: Supported
• Released: 2025-09-21
• Knowledge cutoff: 2025-07-31