The Qwen team at Alibaba Cloud (Qwen is written in Chinese as Tong Yi Qian Wen) develops the Qwen family of models, one of the most prominent open-model efforts to come out of China. Its GitHub organization, QwenLM, describes itself as the home of “Alibaba Cloud’s general-purpose AI models,” with the work attributed to the Qwen team at Alibaba Cloud / Alibaba Group.
The Qwen line spans many model types. Beyond the core large language models, the team’s repositories include Qwen3-VL (multimodal vision-language), Qwen3-Coder (code), Qwen3-Omni (described as supporting “text, audio, images, and video, as well as generating speech in real time”), Qwen-Image (image generation and editing), and Qwen-Agent (an agent framework with function calling and code interpretation). The team also ships qwen-code, “an open-source AI coding agent that lives in your terminal.”
Qwen models are distributed widely on Hugging Face Hub and ModelScope, with many released under the Apache 2.0 license, alongside some source-available and research-only licenses and proprietary variants served through Alibaba Cloud. The companion entry in this library’s models catalog (qwen-models) tracks the specific current flagship.
Why business readers should care: Qwen is a leading non-Western open-model family, and its breadth across text, vision, audio, code, and agents makes it a serious option for organizations seeking capable models they can download, inspect, and self-host.