Hajra Naz
Chinese AI models are rapidly gaining traction in Silicon Valley. They are becoming essential to American companies and earning praise from tech leaders.
Chinese developers such as Alibaba, Z.ai, Moonshot, and MiniMax have gained an edge by offering open-weight language models at much lower costs than U.S. rivals.
This trend highlights the limits of U.S. export controls on advanced chips, which have not stopped Chinese developers from approaching the capabilities of Silicon Valley giants.
Big Companies Embrace Chinese AI
In October, Airbnb CEO Brian Chesky revealed the company chose Alibaba’s Qwen over OpenAI’s ChatGPT, calling it “fast and cheap.”
Social Capital CEO Chamath Palihapitiya said his firm moved much of its work to Moonshot’s Kimi K2 because it was “way more performant” and significantly cheaper than OpenAI and Anthropic models.
Social media users also noted that popular U.S.-developed coding assistants, Composer and Windsurf, appear to be built on Chinese AI models.
Developers Cursor and Cognition AI have not confirmed this, but Z.ai said the reports align with its internal findings.
Chinese AI Becoming the Standard
Nathan Lambert, founder of the Atom Project, said these examples are just the “tip of the iceberg.”
“Chinese open models have become a de facto standard among startups in the U.S.,” Lambert told Al Jazeera.
He added that top American AI startups are training models on Qwen, Kimi, GLM, or DeepSeek, though many are reluctant to disclose their use of Chinese technology.
Usage and Popularity
While precise usage is hard to measure, data shows Chinese AI tools are growing in popularity:
-
MiniMax M2, Z.ai GLM 4.6, and DeepSeek V3.2 were among the top 20 most used AI models last week (OpenRouter data).
-
Four of the top 10 programming models were from Chinese firms.
-
Open-weight Chinese models had over 540 million cumulative downloads as of October (Atom Project/Hugging Face data).
Chinese models appeal especially to early-stage, cost-conscious startups, while high-resource organizations prefer premium U.S. models.
How Chinese Models Work
Unlike U.S. platforms like ChatGPT, Chinese open-weight LLMs make their trained parameters publicly available.
Running them at scale requires large compute power, which providers charge for. Developers like Z.ai and DeepSeek use older-generation chips not covered by U.S. export controls, reducing hardware costs.
“The success of these Chinese models demonstrates the failure of export controls,” said Toby Walsh, AI expert at the University of New South Wales.
“They’ve built better models that are smaller and run on older hardware. Necessity is the mother of invention.”
Chinese models can be offered at much lower prices. AllianceBernstein estimated DeepSeek’s models were up to 40 times cheaper than OpenAI’s last February.
Global Implications
Experts say China’s AI progress has been underestimated. Greg Slabaugh, AI professor at Queen Mary University of London, noted that the rise of open-weight models makes Chinese capabilities globally consumable.
Some analysts compare China’s AI strategy to the solar panel industry, flooding markets with low-cost options.
However, U.S. tech giants still dominate high-end markets and regulated sectors like finance, healthcare, and defense, where security and trust are key.
Rui Ma, founder of Tech Buzz China, predicts AI adoption may follow the Android vs. iPhone trajectory, where affordability drives scale, but high-value margins concentrate at the premium end.
FAQs
1. Which Chinese AI models are gaining popularity in the U.S.?
Models like Alibaba Qwen, Moonshot Kimi, Z.ai GLM, MiniMax M2, and DeepSeek V3.2 are widely used in startups and companies.
2. Why are U.S. companies adopting Chinese AI?
Chinese AI models are faster, cheaper, and open-source, allowing startups to experiment at lower costs.
3. How do open-weight models differ from U.S. AI models?
Open-weight models make their trained parameters publicly available, unlike proprietary models like ChatGPT.
4. Have U.S. export controls affected China’s AI growth?
No. Chinese developers use older-generation chips, bypassing export controls while keeping costs low.
5. What are the long-term implications of Chinese AI adoption?
Chinese AI may dominate affordable AI for startups, while U.S. AI maintains premium markets in regulated sectors.
