Qwen-Image-2.0 improves text and image rendering in compositions for creating infographics and posters

Alibaba has introduced Qwen-Image-2.0the update of its artificial intelligence model for image generation that improves text rendering and the realism of scenes, including people, to offer photorealistic results.

Qwen-Image-2.0 is a model that combines image generation and editing and that supports both text input and other images to offer realistic and professional results. According to Alibaba, it stands out in tests for achieving “superior performance.”

The model improves the following of instructions, and it also does so in composite images that are explained by areas, in which it manages to render the requested elements and maintain the coherence of the whole.

Alibaba has highlighted the precision with which Qwen-Image-2.0 meets complex rendering instructions, thanks to supports indications of up to a thousand tokens, allowing users to make more detailed descriptions.

The results are precise, with professional finishes and, in addition, with aesthetic quality, as explained in the statement shared on the Qwen blog. This is carried over to compositions that include text and image, in which different typographic styles are supported, such as posters, comics and infographics.

The realism It is another point that Alibaba highlights, and not only in the scenes, but also in everything that appears in them, such as people, nature and architecture.

To do this, Qwen-Image-2.0 takes into account elements such as material properties and spatial orientation, and preserves lighting, reflections and perspective. And for photorealism, it naturally recreates actions, expressions, textures and colors.