The model was uploaded in November to an orbital space computing centre operated by Chinese aerospace start-up Adaspace Technology. There, Qwen-3 successfully carried out multiple inference tasks, confirming that large AI models can function reliably in the harsh environment of space. The network is intended to support physical AI applications as well as AI model training and inference directly in orbit, reducing latency and dependence on ground-based infrastructure. Running AI models in space offers several strategic advantages. Alibaba Cloud, the cloud computing and AI arm of Alibaba Group, has positioned Qwen-3 as a general-purpose model capable of handling a wide range of tasks.