An intensive action campaign targeting the network infrastructure of the intelligent era has started with the first shovel of earth in Xiongan New Area. The Cyberspace Administration of China (CAC) has joined forces with the cyberspace administrations of Beijing, Shanghai, Zhejiang, and Shenzhen, along with five leading large model companies, to launch the "Artificial Intelligence Large Model IPv6 Capability Enhancement Special Action." The goal is clear: to promote comprehensive support for generative large model applications on IPv6, and to strengthen the network foundation for the intelligent era.

IPv6 was designed by the Internet Engineering Task Force (IETF) as the next-generation IP protocol to replace IPv4. With 128-bit addresses, its total address pool is 2 to the power of 128, and there's a popular saying that it can assign an address to every grain of sand on Earth. As large model applications emerge rapidly and token traffic grows exponentially, the network's capacity has reached a new threshold. IPv6, with its massive address space, efficient routing mechanism, and inherent security capabilities, has become the key foundation supporting these applications.

This special action will last for one year and includes many practical tasks: enhancing the large model IPv6 application capability on a large scale, improving management standards, raising the IPv6 support level of infrastructure, promoting large model applications to increase traffic, especially raising the proportion of IPv6 access for large model API calls. It will also push forward the integration and innovation of large models with IPv6, driving the Internet toward a "native intelligence" direction.

At the same event, the IPv6 single-stack deployment initiative in Xiongan New Area was also launched. This new city will work collaboratively from four dimensions: laying solid foundations, broad empowerment, strengthening breakthroughs, and cultivating an ecosystem. The goal is set for 2030, aiming to have all newly built areas run IPv6 single-stack on a large scale.