Chinese Open-Weight AI Models Surge as Businesses Shift From US Providers

Daily token volume for Zhipu's GLM-5.2 has surged 50-fold on Vercel since mid-June, as global businesses increasingly shift from premium US closed-source AI systems to cheaper Chinese open-weight alternatives¹. GLM-5.2 operates at roughly one-fifth the cost of Anthropic's Claude Opus 4.8.

The adoption spike is part of a broader pattern documented across multiple outlets. Global businesses are switching from products such as OpenAI's GPT and Anthropic's Claude to Chinese open-weight models that offer near-frontier performance at sharply lower price points. Deep Seek's V4 Flash is also cited as part of the wave. A Bloomberg report frames Deep Seek as championing what it calls China's bid to flood the world with cheap AI³.

Separately, Beijing-based Moonshot AI unveiled Kimi K3, which the company describes as China's largest AI model to date at 2.8 trillion parameters². Moonshot AI said Kimi K3 closes the gap with leading US systems, though it still trails Anthropic's Claude Fable 5 and OpenAI's GPT 5.6 Sol on overall performance. The company said Kimi K3 beat Claude Opus 4.8 and GPT 5.5 on benchmarks including coding and general agents. Kimi K3 consistently outperformed other tested models, according to Moonshot AI.

Moonshot AI was founded in 2023 and counts Alibaba and Tencent among its backers. The company raised $2 billion in May at a valuation exceeding $20 billion.

A separate Bloomberg headline references China's new AI as a factor in a broader tech rout, though details beyond the headline are not available.

Vercel, the San Francisco-based cloud platform for AI web development, reported the GLM-5.2 token-volume surge on Tuesday. ANALYSIS The 50-fold increase on a US-based developer platform indicates that the shift toward Chinese open-weight models is reaching developer ecosystems outside China, not just domestic markets.

ANALYSIS The cost differential — GLM-5.2 operating at about one-fifth the price of Claude Opus 4.8 — creates direct commercial pressure on US model providers at a moment when Anthropic is preparing for an IPO and investing in political engagement around AI regulation.

Moonshot AI's Kimi K3 benchmarks add a capability dimension to the cost story. If Chinese labs can deliver models that match or exceed the second tier of US frontier systems on key benchmarks while pricing aggressively, the competitive dynamics for Anthropic and OpenAI extend beyond their flagship models to the broader product lineup.

The developments arrive as Anthropic schedules investor meetings ahead of a potential IPO later this year ctx and as Anthropic CEO Dario Amodei has made a $1 million personal donation to a super PAC focused on AI safety regulation ctx. Pricing pressure from Chinese open-weight competitors adds a commercial variable to Anthropic's public-market narrative.