AI news ·
Qwen-Image-2.1: Compact, Efficient, and Unified Image Creation
Qwen released a 7-billion-parameter open-source model that combines image generation and editing, supports transparent output and multiple reference inputs, and is available under a research license.
Qwen has released an open-source 7-billion-parameter visual model that combines image generation and editing into a single system. The model supports transparent image output and can accept up to ten reference inputs during operation. Its smaller parameter count is designed to fit local creative workflows that previously required larger, more resource-intensive models.
The release broadens access to unified image creation and editing tools for professionals working across creative, educational, marketing, product development, sales, scientific research, and technical writing fields. The model is distributed under a research license, which means any planned commercial use requires a separate review of the licensing terms.
Source: https://qwen.ai/