China Doesn't Need to Win the AI Race to Break It
The entire debate around Chinese AI has been framed as a race: can Chinese labs catch U.S. frontier models? By most assessments, no, and the frontier gap may actually be widening. But after a recent discussion with Goldman Sachs' Asia internet research team, I'm convinced that's the wrong question entirely. Chinese models don't need to beat U.S. frontier models. They only need to be good enough, at a fraction of the price, to reprice the economics of AI everywhere.
The market is splitting into two tiers. Premium frontier models capture high-value, complex workloads at premium pricing, while low-cost inference captures everything else. And "everything else" is most of AI usage: inbox cleanup, summarization, drafting, and routine white-collar workflows that don't need the best model in the world. Chinese models price around the $0.20 level while top U.S. models run $4-8. When the cost gap is 20-40x and the capability gap doesn't matter for the task, token volume migrates. On some third-party platforms, Chinese models already approach roughly 50% of usage. The playbook is familiar from Chinese e-commerce and EVs: subsidize adoption, undercut on price, win volume, consolidate, monetize later. Negative gross margins today aren't actually a bug, but they're becoming the strategy.
Here's the twist, though: innovation globalizes faster than profits. Many Chinese advances ship open-source or open-weight, including sparse attention, memory reduction, and other inference-efficiency techniques. U.S. labs adopt them quickly, and they already have with cheaper flash-tier models. The value accrues to whoever packages and distributes cheap inference: cloud platforms, application layers, and enterprises. This means Chinese AI can compress global inference pricing without Chinese companies capturing the profit pool. The deflation is exported even when the profits aren't.
There's also a self-improvement engine underneath it all. Once a model attracts real usage, it generates the data to improve its next version. Coding usage matters most here because it provides the highest-quality feedback data there is. This is why export controls and model-access restrictions may slow specific channels but can't stop the flywheel if domestic Chinese adoption gets large enough. Distillation from U.S. models mattered early on, but real user data matters far more from here.
Put these together and you get two starkly different scenarios for markets. In the first, frontier leadership is everything: premium pricing holds, the return on U.S. hyperscaler capex is protected, and the valuation premium on frontier labs is justified. In the second, the clearing price of inference collapses toward good-enough levels, token volume migrates down-tier, and the profit assumptions embedded in U.S. AI valuations, cloud capex, and semiconductor demand all need revisiting. The frontier race decides the first world, while the price of inference decides the second, which is the one nobody is positioned for.
The positioning problem is compounded by a proxy problem. When investors turn bullish on China AI, they buy Alibaba or KWEB, but KWEB is internet and software exposure, not AI. The real Chinese catalysts live in domestic chips, memory, robotics, and compute infrastructure, which is exactly where constrained Nvidia access pushes incremental inference demand. Investors have short memories, and most have already moved on from DeepSeek. That is precisely why a DeepSeek 2.0 moment, likely from hardware rather than chat models, would catch the market using the wrong instruments to express the right view.
In the end, China AI isn't a single-stock question or a national-security footnote, but a global macro variable. The frontier race gets the headlines, but the clearing price of inference is what reprices markets. China doesn't have to win the race to change the economics for everyone running it.