阿里云推出了Qwen-Image-3.0-Pro图像生成模型，该模型支持最多4.5k令牌的输入能力[1]。模型可精确渲染小至10像素的文本[1]，并原生支持12种语言和20多种字体的渲染[1]。
在功能方面，Qwen-Image-3.0-Pro能够处理复杂的布局设计，包括报纸、漫画、菜单等多种复杂格式的生成[1]。该模型还支持前缀完成和函数调用等功能[1]。
API定价采用分层设计[1]：1K图像输入和2K图像输入均为每张图像$0.003，图像输出方面，1K为每张$0.04，2K为每张$0.075[1]。此外，模型的请求频率限制为每分钟1次[1]。
Alibaba Cloud has introduced Qwen-Image-3.0-Pro, an advanced image generation model capable of processing inputs up to 4.5k tokens [1]. The new model supports the generation of complex layouts and formats, including newspapers, comics, and menus, while rendering text as small as 10 pixels with precision [1]. The model also provides native support for 12 languages and over 20 fonts, enabling localized content creation [1].
Pricing for the API is structured based on image resolution: 1K image input costs $0.003 per image and 1K image output costs $0.04 per image [1]. For higher resolution 2K inputs, the cost remains $0.003 per image, while 2K outputs are priced at $0.075 per image [1]. The service operates with a rate limit of 1 request per minute [1].