Where Qwen-Image-2.1 is strongest in Text to Image: closest to the category frontier among all models in Lighting, Knowledge and Reasoning
Our Text to Image taxonomy measures 9 capabilities and 10 use cases, each with its own leaderboard. Capabilities are the individual skills that go into making an image.
Qwen-Image-2.1 is closest to the category frontier in Lighting, followed by Knowledge, Reasoning and Layout. Against Qwen Image 2.0 it closes the gap to the frontier on every capability, with the largest gains in Layout, Lighting and Knowledge.
➤ Lighting covers reflection, refraction, shadows, and caustics.
➤ Knowledge covers real landmarks, species, and domain facts across science and common sense.
➤ Reasoning covers idiom interpretation, maths and science, spatial and temporal reasoning, and concept mixing.
