Qwen Releases Bilingual Open-Source Image Model

Alibaba's Qwen team has released a new text-to-image model, Qwen-Image 2512, expanding its suite of open-source AI tools. The generator is designed to create images from text prompts and is notable for its native support for both English and Chinese languages.

Released under the permissive Apache 2.0 license, Qwen-Image 2512 is available for both commercial and research applications without significant restrictions. This open approach continues Qwen's strategy of contributing foundational models to the community, following its well-regarded releases of large language models.

The model's bilingual capability is a key differentiator in a crowded field of image generators. By performing well with Chinese prompts, Qwen-Image addresses a significant need for tools that can create more culturally and linguistically specific visual content. This could empower developers and artists in Chinese-speaking regions to build more relevant applications. Interested users can find the model and usage details on its official Hugging Face repository.