Chinese open-weight models claim top 5 spots on OpenRouter by usage, threatening US frontier labs
Chinese AI models from Tencent, Xiaomi, DeepSeek, MiniMax and Z.ai now occupy the top five positions by weekly token usage on OpenRouter, a major AI marketplace. The shift threatens to commoditize premium models from OpenAI and Anthropic as businesses reserve expensive frontier systems for only their hardest problems.
Chinese Models Capture Top Usage Spots
Chinese AI models from Tencent, Xiaomi, DeepSeek, MiniMax and Z.ai now occupy the top five positions by weekly token usage on OpenRouter, a major marketplace providing access to hundreds of AI systems. All five are open-weight models that users can download, customize and run on their own infrastructure.
The usage shift comes as Moonshot AI's Kimi K3, released this week, reportedly rivals Anthropic's Fable and OpenAI's GPT-5.6 in key benchmarks—challenging earlier assessments of China's AI capabilities.
Price Gap Drives Enterprise Adoption
Most corporate AI applications do not require frontier-level intelligence, according to industry executives interviewed. Businesses increasingly use cheaper systems for routine coding, summarization, data extraction and customer service, reserving premium models for their most complex problems.
"There are going to be open-source models that eventually handle 95% of enterprise queries, and that remaining 5% may go to OpenAI or Anthropic," one AI investor said.
Kong CEO Augusto Marietti confirmed open-weight usage has surged over the past quarter, citing flagship models as "too expensive." Mozilla CTO Raffi Krikorian said cheaper models can cost up to 50 times less for routine tasks, comparing using frontier AI for everyday work to "driving a Ferrari to Whole Foods."
US Labs Respond
Some American companies are attempting to compete in the open-weight space:
- Thinking Machines, launched by former OpenAI CTO Mira Murati, debuted an open-weight model this week built for deep customization
- Nvidia is expanding its Nemotron family of open models, betting customizable AI will drive chip and software demand
- xAI open-sourced Grok Build, the software behind its coding agent
Shifting Competitive Landscape
AnthropicCEO Dario Amodei stated in May that China remained 6-12 months behind the US in the most dangerous cyber capabilities. Ten weeks later, Moonshot's release suggests that timeline has compressed significantly.
The shift poses risks to OpenAI and Anthropic's planned IPOs, whose valuations depend on frontier AI remaining scarce and commanding premium prices. Both the AI spending boom and stock market performance have become heavily dependent on a small group of companies whose business models assume continued pricing power.
What This Means
The bifurcation of the AI market into premium frontier models and cheap open-weight alternatives represents a fundamental challenge to US labs' business models. If Chinese companies can deliver "good enough" performance at a fraction of the cost while offering customization advantages, frontier labs may struggle to justify the massive capital investments required to maintain their lead. The question is no longer whether Chinese models will catch up, but whether the gap matters when most enterprise workloads don't require cutting-edge capabilities. Mozilla's Krikorian summarized the mood: "They're clearly terrified."
Related Articles
Open Model Race Intensifies: Thinking Machines, Poolside, Moonshot Ship Competing Frontier Releases
A dense wave of open-weight model releases—including Thinking Machines' first model Inkling, Poolside's Laguna S2.1, and Moonshot AI's Kimi K3—signals that consolidation predictions for AI labs have not materialized. Licensing terms, particularly Kimi K3's noncommercial agreement requirement, are emerging as a new front in US-China AI policy.
Open Model Race Intensifies: Thinking Machines, Tencent, Poolside, Moonshot Ship Frontier-Class Releases in Same Week
A wave of open-weight model releases from Thinking Machines, Tencent, Poolside, Moonshot AI, and Meituan signals that model-building capacity is spreading rather than consolidating. The releases range from a 1.6 trillion-parameter MoE trained entirely on Chinese accelerators to a noncommercial-licensed model raising new questions about US-China AI trade.
Moonshot AI Releases Kimi K3 Weights: 2.8 Trillion Parameters, Tighter Commercial License
Moonshot AI has released the weights for Kimi K3, a 2.8 trillion parameter model weighing in at 1.56TB on Hugging Face. The new license drops the 'modified MIT' framing and now requires companies earning over $20 million in 12-month revenue from Model-as-a-Service offerings to sign a separate agreement with Moonshot.
Moonshot AI releases Kimi K3, largest open-weight model at 2.8 trillion parameters
Moonshot AI released Kimi K3 on July 16, 2025, an open-weight model with 2.8 trillion parameters. The model represents the largest openly available model by parameter count, entering what the industry categorizes as the 3T class.
Comments
Loading...