The battle between China and the United States for supremacy in artificial intelligence escalated on Friday when Chinese startup Moonshot AI released a new AI model that appeared to reduce the lead of well-funded American competitors.
Moonshot said the Kimi K3 model is the world’s largest open source AI system, allowing anyone to freely use, modify and build on it. The company said Kimi K3 performs better than OpenAI and Anthropic’s models in some key tasks.
The release coincided with an address by Xi Jinping, China’s leader, in which he outlined an ambitious vision for global AI development that made China a champion of an open approach to the technology.
“AI development should not be a solo effort by a single country, but a symphony of international cooperation,” Mr. Xi said. China is “willing to work with all parties to seize the opportunities of AI development,” he said.
The market reaction was immediate. Stocks fell on Friday, with the Nasdaq falling about 1 percent, as investors sold shares of U.S. companies that make the computer chips needed for AI, including Nvidia and Intel.
The developments highlighted the geopolitical stakes of the US-China competition for leadership in what many consider to be the defining technological race of our time. President Trump has repeatedly described maintaining America’s leadership in artificial intelligence as a strategic imperative to maintaining the country’s economic and national security.
The emergence of AI as an enabling technology with far-reaching economic and military implications has led the United States to slow China’s progress by restricting its access to the advanced chips needed to train cutting-edge AI systems.
The release of a free, open-source Chinese AI model that could rival the performance of costly, compute-intensive systems from Silicon Valley’s deepest companies has renewed concerns about whether the data center industry’s massive spending spree is justified.
Over the past 18 months, Chinese tech companies have released a steady stream of open-source AI models that match the performance of their American competitors while claiming to require fewer computing resources. This combination of strong performance and lower cost has made Chinese AI systems the preferred choice for many developers worldwide.
Moonshot’s latest release adds to this momentum, raising the prospect that a Chinese company could soon rival or even surpass the cutting-edge models of leading American AI companies, which have spent hundreds of billions of dollars building closed and expensive-to-operate models.
According to benchmarks from Vals AI, an independent company that evaluates the performance of AI models, Kimi K3 performs just below Anthropic’s Fable 5 model while outperforming OpenAI’s flagship GPT-5.6 Sol model, said Rayan Krishnan, CEO of Vals AI.
Anthropic’s latest AI models have become so powerful that their capabilities – including identifying vulnerabilities in critical infrastructure such as banking networks and power grids – have alarmed governments around the world.
Graham Webster, a Stanford University professor who studies China’s open-source AI ecosystem, said it doesn’t matter much whether a model is state-of-the-art because the performance gap between leading American and Chinese models has narrowed.
“Everyone has a different number about how many months China is behind the U.S., and the estimates are around six months,” he said. “That’s not much of a clue.”
Moonshot announced that it will release full details about Kimi K3 at the end of the month.
Samm Sacks, a senior fellow at the Johns Hopkins School of Advanced International Studies who attended the AI conference in Shanghai on Friday where Mr. Xi spoke, said many experts believe it is only a matter of time before a Chinese company releases a freely available model to match Anthropic’s most advanced systems. Such a breakthrough, Ms. Sacks said, would raise difficult questions for regulators trying to rein in powerful AI technology after its release.
Last month, another Chinese startup, Z.ai, released a model called GLM-5.2 that was almost as powerful as Anthropic’s Fable 5 model, the leading American system. And unlike Fable, this Chinese model was widespread.
Many software developers and startups in Silicon Valley quickly adopted the model, primarily because it was much cheaper than the leading American systems. It came just as US companies realized they needed to find ways to reduce their AI spending.
According to Vals AI benchmarks, Kimi K3 is even more powerful than GLM-5.2. The model comes at a time when the Trump administration is considering regulating the technology. Many Silicon Valley executives have begun to support such regulation, saying leading technologies can enable malicious cyberattacks and potentially help build biological weapons.
Moonshot’s AI has been widely used in China since the Beijing-based startup was founded in 2023.
Almost all of China’s leading AI systems are open source, helping them gain users around the world. But unlike the closed systems popular in Silicon Valley, the Chinese models proved harder to profit from. The challenge of leveraging global adoption to build a profitable business is now fracturing some of China’s leading AI research labs, including internet giant Alibaba.
Chinese AI researchers regularly cite a lack of computing power – caused by US export controls – as the biggest obstacle facing the industry, and companies have spent heavily to secure scarce chips. Some companies, including MiniMax and Z.ai, have gone public to raise money from investors. Alibaba is increasingly steering customers toward models where users have to pay.
But the scale of fundraising by Chinese AI firms pales in comparison to their U.S. competitors. In May, Moonshot raised $2 billion in a funding round backed by China Mobile, the state-owned telecommunications company, and the investment arm of food delivery giant Meituan. In comparison, Anthropic raised $65 billion in the same month.
Moonshot has instead focused on making the most of its limited computing resources. Speaking at the World Economic Forum in Davos this year, the company’s president, Yutong Zhang, said efficiency had been a necessity from the start.
“We knew we didn’t have the luxury of simply increasing computing power,” she said.
Anthropic and OpenAI have accused Chinese companies of improperly collecting data from their AI systems to speed up the development of competing models. However, experts say Moonshot and Z.ai’s recent advances cannot be explained by distillation alone. Years of operating under export controls have forced Chinese companies to innovate within strict computing restrictions.
“The Chinese models are excellent for more than just distillation,” said Mr. Webster of Stanford. “There is real innovation.”
Xinyun Wu contributed research from Taipei, Taiwan.
https://www.nytimes.com/2026/07/17/business/china-ai-moonshot-kimi.html
