← All models

Qwen VLo

Published Publisher Alibaba

Model Overview

Qwen VLo is a unified multimodal understanding and generation model that combines image understanding, generation, and editing capabilities. The official preview was released on June 26, 2025, and was accessible through Qwen Chat at the time. (qwenlm.github.io)

Main Capabilities and Uses

  • Generate images from text, including general images and posters in Chinese and English.
  • Edit uploaded images based on natural-language instructions, such as replacing backgrounds, adding objects, changing styles, or making multiple modifications.
  • Generate visual annotations such as detection, segmentation, and edge information, and re-analyze generated content.
  • Use dynamic resolution and progressive image generation for scenarios such as posters, illustrations, website banners, and social media covers. (qwenlm.github.io)

Limitations

At release, the official announcement stated that the model was still in preview and that results might be factually inaccurate, not fully consistent with the original image, or fail to follow instructions; recognition and understanding of images and user intent could also be unstable. Multiple-image input and generation of images with extreme aspect ratios had not been officially launched at release, and their current availability has not been verified. (qwenlm.github.io)

Pricing

As of November 18, 2026, checks of the Alibaba Cloud China site model pricing and Alibaba Cloud international site model pricing did not verify any official pricing applicable to this preview version; the billing methods and amounts for input and output have not been verified. Prices for other Qwen image models cannot be directly treated as prices for this model. (help.aliyun.com)

Official Source