Chinese companies have released advanced AI models that rival US offerings, yet the latest ones sparked little of the earlier market panic. DeepSeek's R1 in January 2025 triggered a trillion-dollar selloff in US tech stocks, but Z.ai's GLM-5.2 last month did not. The shift reflects China's growing role in open-source AI.
When DeepSeek released its open-source R1 model in January 2025, it was reported to match leading US systems while remaining free to download. US lawmakers quickly proposed bans on government devices.
Z.ai followed with GLM-5.2 last month. The model ranked as the top open-source AI by Artificial Analysis and placed fifth in usage on OpenRouter. Seven of the platform's top ten models now come from Chinese firms.
Experts note the contrasting approaches. Chinese developers release models for local use at no cost. Most US firms keep models in the cloud and charge for access. Serge Belongie of the University of Copenhagen said Chinese firms are attempting to pressure Western labs through aggressive open-source releases.
Mikhail Belkin at the University of California, San Diego, observed that the Chinese models match US performance from six to nine months ago and offer greater stability. Philip Torr of the University of Oxford warned that US export controls could push China to build independent ecosystems.
