China records 700+ filed generative AI models under ‘AI+’ push

Beijing’s disclosure that more than 700 generative AI large-model products have completed filing is a clear signal of just how dense China’s model ecosystem has become. Combined with metrics like an expanded digital economy share of GDP and rapid growth in network infrastructure, this update shows that frontier models such as Tongyi Qianwen and DeepSeek now sit inside a broad, state-orchestrated industrial system rather than as isolated tech experiments.

For the race to AGI, the filing number matters less than what it implies: China is normalizing generative AI as regulated infrastructure. Registration and oversight bake frontier models into long-term planning for sectors from manufacturing to rural e‑commerce, while also giving the state visibility into which architectures and providers are gaining traction. That creates a different competitive landscape from the U.S., where a handful of labs dominate but regulation is still piecemeal. If China can turn sheer model proliferation plus massive HarmonyOS device penetration into robust agent ecosystems and embodied deployments, it could quickly close the perceived capability gap with Western labs, especially in applied, domain-specific intelligence.