Particle.news
Download on the App Store
5 ARTICLES
·
2mo ago

Alibaba Unveils Qwen-VLo Preview With Multimodal Image Generation and Editing

Available as a free preview on Alibaba’s chat interface, the model showcases progressive left-to-right rendering, inline image edits, multilingual prompts, dynamic resolution training, upcoming multi-image input.

While giving instructions on Qwen VLo the user will be free to write in multiple languages, including in Chinese and English.
The Qwen VLo is also capable of image annotation tasks, such as edge detection and segmentation
Image
Alibaba has launched Qwen VLo, a new AI model for generating and editing images using text prompts.

Overview

  • Qwen-VLo is accessible without login on Alibaba’s chat interface in preview form, marking the latest open-source release in the Qwen AI series.
  • The model constructs images progressively from left to right and top to bottom to deliver finer control over generation.
  • It supports both text-to-image and image-to-image workflows with prompts in multiple languages, including English and Chinese.
  • Inline editing lets users adjust or transform generated and input images without disrupting their structural integrity.
  • Dynamic resolution training enables varied aspect ratios and multi-image input is rolling out, though users may encounter inconsistencies during this preview phase.