Foreign media: While Chinese AI startups have rapidly narrowed the technological gap with their U.S. counterparts—Moonshot’s Kimi K3 has become the world’s largest open-source weight model, and China’s share of model token traffic has surged from 1.2% in 2024 to over half by summer 2026, with the performance gap between top-tier models shrinking from seven months to four—the funding bottleneck is becoming increasingly severe.
Between 2023 and 2026, U.S. AI firms attracted over $380 billion in venture capital, while China received only about one-tenth of that amount; in the first quarter of 2026, China’s total VC funding reached just $20 billion, far below the U.S. figure of $267 billion. Compounding the challenge are rising internal inflationary pressures within the AI economy—memory prices continue to climb, and AI talent compensation has increased twelvefold—alongside the fact that enterprise software remains largely focused on domestic markets and operates under extended profitability cycles, making it difficult for Chinese startups to replicate the private fundraising trajectories of OpenAI (over $100 billion) or Anthropic ($65 billion).
As a result, these companies are forced to pursue Hong Kong listings earlier than planned—for example, Zhipu AI and MiniMax raised only $558 million and $620 million, respectively—and must explore alternative financing channels such as private credit, all while maintaining competitiveness with funding levels roughly one-tenth those of their American peers.
Original source: toutiao.com/article/1878459303494656/
Disclaimer: The views expressed in this article are solely those of the author.