Chinese AI models are popular around the world and beat US competitors in some areas

China’s efforts to dominate the world of artificial intelligence may be paying off, as industry insiders and tech analysts tell CNBC that Chinese AI models are becoming more popular and are matching — and even surpassing — those from the US in terms of performance.

AI has become the latest battleground between the US and China, with both sides viewing it as a strategic technology. Washington continues to restrict China’s access to advanced chips designed to power artificial intelligence amid fears the technology could threaten US national security.

It has led China to pursue its own way of increasing the appeal and effectiveness of its AI models, including relying on open-source technology and building super-fast software and chips.

Like some of the leading US firms in the space, Chinese AI firms are developing so-called large-scale linguistic models, or LLMs, which are trained on large amounts of data and support applications such as chatbots.

Unlike the OpenAI models that power the more popular ChatGPT, however, many of these Chinese companies create open source LLMs, or open weight, that developers can download and build on for free and without strict licensing requirements from the developer.

On Hugging Face, an LLM repository, Chinese LLMs are the most downloaded, according to Tiezhen Wang, a machine learning engineer at the company. Qwen, a family of AI models created by a Chinese e-commerce giant Alibabamost famous for Hugging Face, he said.

“Qwen is quickly gaining popularity due to its high performance in competitive environments,” Wang told CNBC via email.

He added that Qwen has an “excellent licensing model” that means it can be used by companies without the need for “extensive legal review.”

Qwen comes in various sizes, or parameters, as they are known in the world of LLMs. Large parameter models are more powerful but have higher computational costs, while smaller ones are cheaper to run.

“Regardless of which model you choose, Qwen is likely to be one of the best performing models available right now,” added Wang.

DeepSeek, a startup, has also made waves recently with a model called the DeepSeek-R1. DeepSeek said last month that its R1 model competes with OpenAI’s o1 – a model designed to think or solve complex tasks.

These companies claim that their models can compete with other similar open sources MetaLlama, and closed LLMs like those from OpenAI, across a wide range of functions.

“Over the past year, we’ve seen an increase in Chinese open source contributions to AI with really strong performance, low operating costs and high performance,” Grace Isford, a partner at Lux Capital, told CNBC in an email.

Open-sourcing technology serves many purposes, including innovation as more developers have access to it, and building a community around the product.

It is not only Chinese companies that have introduced open source LLMs. Facebook parent Meta, and European start-up Mistral, also have open-source versions of AI models.

But with the technology sector caught in the crosshairs of a world war between Washington and Beijing, open-source LLMs offer Chinese firms another advantage: making their models globally applicable.

“Chinese companies would like to see their models applied outside of China, so this is a sure way for companies to become global players in the AI ​​space,” Paul Triolo, a partner at global consulting firm DGA Group, told CNBC in an email.

While the focus is on AI models right now, there is also debate about which applications will be built on top of them – and who will dominate this global internet platform going forward.

“If you think of these frontier AI models as table stakes, it’s about what these models are used for, like accelerating frontier science and engineering technology,” Lux Capital’s Isford said.

Today’s AI models have been compared to applications, such as Microsoft Windows, GoogleAndroid and an appleiOS, which has the potential to dominate the market, as these companies do in mobile phones and PCs.

If true, this makes the stakes for building an outstanding LLM high.

“See [Chinese companies] they see LLMs as the biosphere of the future,” Xin Sun, senior lecturer in Chinese and East Asian business at King’s College London, told CNBC in an email.

“Their future business models will depend on developers who join their ecosystem, develop new applications based on LLMs, and attract users and data where profit can be generated later in various ways, including but more than directing users to use their cloud services,” added the sun.

AI models are trained on a large amount of data, which requires a large amount of computing power. Currently, Nvidia leading designer of the chips needed for this, known as graphics processing units (GPUs).

Most of the leading AI companies train their systems on Nvidia’s high-performance chips – but not in China.

Over the past year or so, the US has tightened export restrictions on advanced semiconductor and chipmaking equipment to China. It means NvidiaThe leading chips cannot be exported to the country and the company has had to create semiconductors that comply with sanctions for export.

Despite these limitations, however, Chinese firms are still able to introduce advanced AI models.

“The big Chinese technology platforms currently have enough access to computing power to continue developing models. This is because they have collected a large number of Nvidia GPUs and also use domestic GPUs from Huawei and other industries,” said Triolo of the DGA Group.

Indeed, Chinese companies have been developing efforts to create viable alternatives to Nvidia. Huawei has been one of the leading players in pursuing this goal in China, while firms are interested Baidu and Alibaba has also been investing in semiconductor manufacturing.

“However, the gap in terms of advanced hardware computing will widen over time, especially next year as Nvidia releases its Blackwell-based systems that are banned from being shipped to China,” Triolo said.

Lux Capital’s Isford noted that China has “systematically invested and is expanding its domestic AI infrastructure stack outside of Nvidia with high-performance AI chips from companies like Baidu.”

“Whether or not Nvidia chips are banned in China will not prevent China from investing and building infrastructure to build and train AI models,” he added.