← signals
2026-07-07·ANTHROPIC·chinese model adoption surge
meddown

Chinese AI models have rapidly gained traction among US companies, drawing over 30% of token usage via OpenRouter since...

Chinese AI models have rapidly gained traction among US companies, drawing over 30% of token usage via OpenRouter since February 2026 and peaking at 46%, up from 11% over the prior 12 months (Source 3: CNBC).

window 60devidence 32confidence score 100

confidence score

Strong evidence: 18 independent source classes support this read.

100
medium confidence18 independent source classesothernewscommunitymarketpasses publish gate

signal brief

Chinese AI models have rapidly gained traction among US companies, drawing over 30% of token usage via OpenRouter since February 2026 and peaking at 46%, up from 11% over the prior 12 months (Source 3: CNBC). This surge is driven by narrowing performance gaps and significantly lower costs—Chinese models are 60-90% cheaper than leading Anthropic and OpenAI models (Source 33). For example, AI startup Lindy moved 100% of its traffic from Anthropic's Claude to DeepSeek, saving millions within months. Concurrently, Anthropic faces competitive pressure from multiple fronts: it is offering token credits and promotions to retain enterprise customers (Source 4: WSJ), while also contending with geopolitical friction—Alibaba banned Anthropic AI for employees after a distillation accusation (Source 34), and a secret tracker controversy eroded trust (Source 35). Despite Anthropic's rapid revenue growth to a $47B run-rate (Source 38), the sustained cost-driven shift toward Chinese alternatives poses a material risk to its US market share and pricing power. The trend shows no sign of reversing as Chinese labs continue launching competitive models like Z.ai's GLM 5.2.

What the sources said:

  • CNBC (Source 3): "Chinese AI models have drawn 30%+ of token use by US companies each week since February 8, peaking at 46%."
  • CNBC (Source 33): "Open source Chinese models can be 60% to 90% cheaper than the leading Anthropic and OpenAI models."
  • WSJ (Source 4): "OpenAI, Anthropic, and other top AI labs are offering startups token credits and special promotions..."
  • Ars Technica (Source 35): "Anthropic quickly removed a tracker secretly monitoring Claude Code users in China after a security researcher exposed the hidden code..."

source data used

Decision support, not stock advice. This signal is research with cited evidence — not a recommendation to buy, sell, or hold any security.