Lenovo launched its Wanquan Hybrid Token Factory brand on November 10 in Zhangjiakou, Hebei, and established its first facility at the local Beijing-Tianjin-Hebei computing power hub node.

The company aims to commercialize AI token production and computing power scheduling.

Lenovo said the product sits in the middle layer of its three-tier architecture. Below it is Wentian, Lenovo’s AI infrastructure covering servers, storage, networks, and heterogeneous computing platforms.

Above it are Qingtian agent services for large government and enterprise customers, and Baiying services for small and medium-sized businesses.

Lenovo executives said the system combines private token pools with public tokens, open-weight models with closed-source commercial models, and different types of chips. Its model routing sends complex tasks to larger models and simpler interactions to smaller ones to improve token production efficiency.

The launch comes as Lenovo seeks to meet rising demand in China for AI inference capacity and related workload management.

🔗 Source: Security Times