Qwen-Image-3.0

On 21 July, Ali released the latest basic model for the generation of images, Qwen-Image-3.0, which is also the third-generation basic model for the generation and editing of the model series. The new model supports a maximum of 4.5 k token super-long input, which can generate, at one time, knowledge maps, complex UI interfaces that cover the integration of elements such as formulae symbols, geometric graphics and logical extrapolation steps, while supporting original rendering of 12 national languages and more than 20 fonts. Kim Xian