Qwen-Image is a new 20B open-source image foundation model by the Qwen team. It excels at complex text rendering (especially Chinese) and precise image editing, while also delivering strong general image generation. Available now in Qwen Chat.
Qwen-Image is a 20B MMDiT image foundation model that achieves significant advances in complex text rendering and precise image editing. It feels like the Qwen team's take on a "4o" moment—a powerful image model integrated directly into Qwen Chat. You can use it just by toggling the "Image Generation" button.
The quality of images this model generates is really impressive. I've found its performance on classical scenes to be genuinely stunning.
What really sets it apart is its incredible ability with text rendering, handling complex Chinese and English text within images better than many open models I've seen. It's a fantastic new tool for creators.
Hi everyone!
Qwen-Image is a 20B MMDiT image foundation model that achieves significant advances in complex text rendering and precise image editing. It feels like the Qwen team's take on a "4o" moment—a powerful image model integrated directly into Qwen Chat. You can use it just by toggling the "Image Generation" button.
The quality of images this model generates is really impressive. I've found its performance on classical scenes to be genuinely stunning.
What really sets it apart is its incredible ability with text rendering, handling complex Chinese and English text within images better than many open models I've seen. It's a fantastic new tool for creators.
Gave it a quick spin. Results look promising and comparable to Flux Kontext
Aiarty Video Enhancer
Big congrats on the launch - it looks seriously impressive!