Conceptual diagram of AI Photo: VCG

Amid growing market attention to the economic effects generated by tokens, "token," a key concept in the field of artificial intelligence (AI), has driven up related service prices as the nation's demand for computing power rises. Notably, major domestic cloud providers increased token service fees by about 30 percent within a 10-day period in March, China Media Group (CMG) reported on Saturday.

Industry observers said that computing power is shifting from a "cost item" for companies to a "strategic resource" that determines competitiveness, with rising prices only a surface phenomenon, while the real change lies in the industry redefining the value of computing power.

This month, China's average daily token usage exceeded 140 trillion, a more than 1,000-fold increase from 100 billion at the beginning of 2024 and up from 100 trillion at the end of 2025, Liu Liehong, head of the National Data Administration, said at a press conference on March 24.

Zhou Hongyi, founder and chairman of 360 Security Technology, told the Global Times in a statement on Sunday that the token economy is a value distribution system centered on tokens, anchored in a six-factor model encompassing electricity, computing power, intelligence, human capital, security capacity and productivity.

A token-driven economy centers on fostering new quality productive forces and will become a core benchmark in the AI era, with its consumption, output efficiency and cost control serving as key indicators of AI capabilities at the corporate, industry and national levels.

Recent rapid advances by Chinese companies in fields such as AI are reshaping public perception, with the surge in token usage reflecting the nation's fast-paced iteration and innovation in AI.

"The substantial increase in average daily Token usage clearly shows that China's AI development has entered a phase of rapid growth," Liu said, noting that AI applications are deepening from conversational functions to decision-making and execution capabilities. The growing focus on "Token going global" also reflects the strengthening competitiveness of China's AI industry.

Liu added that, through joint efforts, China has made phased progress in building high-quality datasets. As of the end of 2025, more than 100,000 such datasets had been established nationwide, totaling more than 890 petabytes—roughly 310 times the size of the digital resources held by the National Library of China, according to Liu.

Notably, on social media platforms, nearly 600,000 users participated in discussions related to the term "token" between March 1 and March 20, CMG reported.

After recently being standardized in Chinese as ciyuan, the term "token" has gained broader visibility among Chinese users, attracting more than 10 million cumulative views on domestic social media platforms.

"From an industry chain perspective, upstream hardware makers such as chip and server manufacturers are the first to benefit from price increases, which are then passed on to end users. This, in turn, forces companies to either improve efficiency or reduce costs by switching to domestically developed computing power. As a result, the current round of price hikes may accelerate the substitution of domestic computing capacity and drive further technological innovation across the industry," said Ruan Qingsong, a professor at the Department of Economics and Finance of Tongji University, according to CMG.

At the recently concluded Zhongguancun Forum 2026, "token" emerged as one of the most frequently discussed buzzwords among entrepreneurs and experts. An industry insider told the Global Times on the sidelines of the forum that the explosive growth in token usage reflects how AI is entering everyday life and empowering various sectors. 

Chinese authorities are stepping up efforts to empower the innovative development of AI and support the building of high-quality datasets.

The National Data Administration will continue to advance data-driven AI innovation, working with all stakeholders to further implement a new round of high-quality dataset development initiatives. Guided by real-world application needs, it will accelerate pilot programs and early trials to build AI-ready, high-quality datasets that are technically feasible, practical and quality-assured, thereby improving both the scale and quality of data supply, according to Liu.