Qwen is Alibaba’s family of AI models, developed by the Qwen Team and spanning language, vision-language, image, and translation tasks. The models are published with links to Hugging Face and GitHub repositories, reflecting a broadly open release approach.
As of June 2026, the Qwen blog names recent models including the Qwen3 foundation models, a Qwen3Guard safety model (posted September 23, 2025), the Qwen-Image and Qwen-Image-Edit image models (posted August 2025), the Qwen2.5-VL vision-language model, and a Qwen-MT translation model. The blog publishes dated posts for each release, and the lineup expands quickly across modalities.
Distribution centers on public model releases via Hugging Face and GitHub alongside Alibaba’s cloud services.
Why business readers should care: Qwen is one of the most widely downloaded open model families globally, making it a practical, low-cost foundation that many third-party AI products quietly build on.