Chinese Open-Source LLMs Capture One-Third of Global Usage Amid US Chip Bans
Updated
Updated · Fortune · Jun 16
Chinese Open-Source LLMs Capture One-Third of Global Usage Amid US Chip Bans
1 articles · Updated · Fortune · Jun 16
Summary
Qwen, MiniMax and DeepSeek now account for one-third of global LLM usage, up from almost zero in late 2024, as developers from Silicon Valley to Africa and Southeast Asia adopt cheaper, open models.
Chinese firms reached that scale by optimizing for efficiency under constraints—US chip bans, tighter capital and weaker domestic software demand—rather than matching the US frontier-model spending race.
DeepSeek became the clearest example after claiming a January 2025 reasoning model was trained for $6 million; by late 2025, its parent said 1 million units of output cost about 3 RMB, roughly one-twentieth of ChatGPT then.
That cost discipline is reinforced by lower input prices—AI engineers earn about 402,000 RMB a year, some provinces halve electricity costs for chip-using facilities, and China produces 1.5 to 2 times as many AI-relevant PhDs as the US.
The result is a parallel AI ecosystem: US companies keep pursuing capital-heavy AGI and agentic systems, while Chinese players use open-source distribution and enterprise applications to spread influence globally.
As Chinese AI becomes 20x cheaper, can America's billion-dollar strategy still win the global technology race?
The AI race is shifting from language to 'world models'. Who is positioned to dominate this next critical frontier?
The Global Rise of Chinese LLMs: 61% Market Share, Open-Source Strategy, and the Future of AI Governance
Overview
Chinese Large Language Models (LLMs) are quickly rising to global prominence by strategically adopting open-source models. This approach allows Chinese AI firms to attract a broad international user base and establish a strong foothold in various generative AI categories. By proactively rolling out open-source models and making them freely available, these companies significantly lower barriers to entry for developers and businesses. This accessibility leads to the creation of diverse AI systems at reduced costs, driving widespread adoption. As a result, Chinese firms are building brand equity and generating positive word-of-mouth within the global AI community.