China is rapidly positioning itself at the forefront of the global open-source artificial intelligence (AI) race, leveraging technological innovation and industrial integration to strengthen its digital economy and global competitiveness.
Chinese artificial intelligence company Moonshot AI has recently made its latest flagship large language model, Kimi K3, fully open-source for public use. With 2.8 trillion parameters, Kimi K3 is now the world’s largest open-source large language model (LLM), marking the first time a flagship AI model of such scale has been released entirely under an open-source framework.
China’s push toward open-source AI has accelerated significantly this year. In just six months, several of the country’s leading technology companies have made their flagship foundation models publicly available, reflecting a broader national strategy to promote collaborative AI development.
According to a report released this year, Chinese-developed open-source models account for 41 percent of total global downloads. The top six most widely used open-source models worldwide are all developed in China. Moreover, for nine of the past 12 months, China has held the record for the open-source model with the largest number of parameters.
China’s open-source AI ecosystem has also become increasingly diverse in scale and capability. Developers have introduced models ranging from hundreds of millions to hundreds of billions of parameters, serving a wide variety of computing needs.
Smaller models can operate offline on smartphones, smart devices and industrial equipment, while larger models support advanced scientific research, software engineering and complex analytical tasks. As a result, AI solutions are now available for users ranging from individual consumers to large industrial enterprises.
The application of China’s open-source models is expanding rapidly across industries. Beyond general-purpose functions, specialized large language models have been developed for sectors such as automotive manufacturing, machinery, electric power and healthcare. These models integrate industry expertise and professional knowledge into AI systems, helping businesses overcome practical implementation challenges and bridge the “last mile” between technological innovation and industrial deployment.
Driven by continuous breakthroughs in core technologies, China’s open-source large language models have become one of the world’s leading forces in the global open-source AI ecosystem. Statistics show that cumulative downloads of Chinese open-source models have exceeded 10 billion, the highest figure worldwide. Today, six of the ten most-downloaded large language models globally are developed in China.
Artificial intelligence is also moving rapidly beyond research laboratories into the real economy. AI technologies are being deeply integrated into manufacturing, energy, transportation, healthcare and other key industries, accelerating industrial upgrading while lowering operating costs and improving productivity.
Open-source development has significantly reduced barriers to AI adoption across industries. Consequently, China’s large language models are no longer limited to conversational AI or question-and-answer applications. They have evolved into practical tools capable of executing real-world tasks, enhancing productivity and improving operational efficiency across manufacturing, energy, transportation, finance and other sectors of the real economy.
According to available data, China now holds nearly 60 percent of the world’s AI patents, underscoring its growing technological leadership. Figures released by the Ministry of Industry and Information Technology (MIIT) indicate that by 2025, the country’s core AI industry had surpassed 1.2 trillion yuan in value, while the number of AI enterprises nationwide had exceeded 6,200.
Experts believe that as the open-source ecosystem continues to expand, artificial intelligence will become increasingly embedded throughout every layer of the real economy. This momentum is expected to accelerate new industrialization, support high-quality growth of the digital economy and further reinforce China’s leadership in global technological competition.
Source: China Media Group (CMG).