🇺🇸▼
The past year marked a turning point for artificial intelligence in China. Since DeepSeek released its R1 reasoning model in early 2025, Chinese companies have consistently delivered AI models that match top Western systems while costing a fraction of the price. Just recently, the Chinese firm Moonshot AI released its latest open-weight model, Kimi K2.5. It came close to proprietary systems like Anthropic's Claude 3.5 Sonnet on various benchmarks, yet costs about one-seventh of Claude 3.5 Opus's price.
On Hugging Face, Alibaba's Qwen family has become one of the most downloaded model series in 2024 and 2025, challenging Meta's Llama models in total downloads. A recent study from the Massachusetts Institute of Technology found that Chinese open-source models have surged in adoption, rivaling US models in total downloads. For developers worldwide, access to top AI capabilities has never been so broad or affordable.
These models differ fundamentally from most US models like ChatGPT or Claude. Users typically pay to access those systems and cannot inspect their internal workings. In contrast, Chinese companies publish their models' weights, which are the numerical values set during training. This means anyone can download, run, study, and modify them. If these open-source AI models continue to improve, they will not only be the cheapest option for top-tier AI; they will also change where innovation happens and who sets the industry standards.
When DeepSeek launched R1, the initial shock focused on its origin. Suddenly, a Chinese team had released a reasoning model that could stand with the best from US labs. However, DeepSeek's lasting impact had less to do with nationality than with distribution. R1 was released as an open-weight model under a permissive MIT license, allowing anyone to download, inspect, and deploy it. DeepSeek also published a paper detailing its training process. For developers using an API, DeepSeek undercut competitors on price, offering access for a fraction of the cost of OpenAI's o1, the leading proprietary reasoning model at the time.
Within days, DeepSeek replaced ChatGPT as the most downloaded free app in the US App Store. This moment spilled into financial markets, triggering a sharp sell-off in US tech stocks that briefly erased about $1 trillion in market value. Almost overnight, DeepSeek went from a little-known team to the most visible symbol of China's push for open-source AI.
China's decision to focus on open source is not surprising. It possesses the world's second-largest concentration of AI talent and a vast tech industry. After ChatGPT became mainstream, China's AI sector underwent a reckoning and emerged determined to catch up. Pursuing open source was seen as the fastest way to close the gap by rallying developers and spreading adoption.
DeepSeek's success gave confidence to an industry used to following global standards. "Thirty years ago, no Chinese person would believe they could be at the center of global innovation," says Alex Chenglin Wu, CEO of Atoms, an AI agent company and a contributor to China's open-source ecosystem. "DeepSeek shows that with solid technical talent, a supportive environment, and the right culture, it's possible to do truly world-class work."