Editor's note: Artificial intelligence is developing rapidly. At the end of March, the National Data Administration officially renamed "Token" as "CiYuan", and "CiYuan economy" immediately sparked heated discussion in mainland China. Data shows that in early 2024, China's average daily CiYuan usage was only 100 billion; by the end of 2025, it had surged to 100 trillion. This March, it even surpassed 140 trillion. With growth of more than a thousand times over two years, does this mean that a new era of intelligent economy has officially arrived?
【Anchor】Artificial intelligence is developing rapidly. At the end of March, the National Data Administration officially renamed "Token" as "CiYuan", and "CiYuan economy" immediately sparked heated discussion in mainland China. Data shows that in early 2024, China's average daily CiYuan usage was only 100 billion; by the end of 2025, it had surged to 100 trillion. This March, it even surpassed 140 trillion. With growth of more than a thousand times over two years, does this mean that a new era of intelligent economy has officially arrived? Welcome to DDN Business Insider.
Today, we are especially joined by three distinguished guests to provide in-depth analysis of the CiYuan economy and future industry opportunities:
Mr. Jiang Han, Senior Research Fellow at Pangoal Institution; Mr. Gu Shuangfei, President of the Nanhua Research Institute; and Mr. Wang Bin, Co-founder of Equal Ocean and Vice Secretary-General of the Industry Innovation Cluster at the China Association for the Promotion of Industrial Development. Hello, everyone!
The concept of Token is not actually a brand-new idea, but earlier this year, it attracted much broader attention. Mr. Jiang, what does this indicate? Is its sudden boom closely related to the buzz around "Lobster" (OpenClaw-typed) AI agents?
【Jiang】My understanding is that the surge in the Token concept is not an isolated event. Judging from the development stages of the AI industry—from technological breakthroughs to the large-scale rollout of scaled applications—this is an inevitable outcome. First, the spike in Token indicates a significant drop in AI inference costs and the explosive expansion of application scenarios. Second, this phenomenon is undeniably closely related to the rise in popularity of core AI agents such as "Lobster." AI agents have the ability to autonomously plan and call tools. The CiYuan consumption for a single agent task can be tens of, to even hundreds of, times that of a simple conversation. This sends a clear signal that AI productivity is taking shape. Token is no longer merely a technical parameter; it becomes the core metric for measuring AI service output and value. This means AI is moving from the "try-it-out" and experimentation phase in the past into a "must-have" stage—meaning the market has already completed early-stage education.
【Anchor】Alright. At this time, the National Data Administration officially renamed Token as "CiYuan". Mr. Jiang, in terms of policy and industry, what does it signify?
【Jiang】The National Data Administration's official renaming of Token as "CiYuan" is a key milestone in the intelligent economy's shift from rough, unstructured growth to standardized, rule-based development. First, this establishes a "value anchor" for the intelligent era. By standardizing token units (CiYuan), they are endowed with the kind of attributes that make "kilowatt-hours" a legal unit of measurement—providing a unified language system for industry statistics, transaction volumes, and settlement.
Second, it also marks a deeper step in the factorization and valorization of data. As a settlement unit that connects technological supply with commercial demand, the official name of CiYuan implies national recognition of its significant role as a core productive factor. This is also a strategic move to seize global discourse power in the intelligent economy. By using a Chinese term, it represents—at least to a certain extent—an early move in rule-making for the AI economy. This can help promote the internationalization of domestic standards.
【Anchor】Yes. Also, I'd like to ask Mr. Wang, what is the significance of this move for the industry and the market?
【Wang】Actually, for the country, the emergence of the concept of "CiYuan" is about establishing rules and setting standards. It is, more than anything, about our voice and discourse power. In this international strategic competitive landscape, we have defined our own original discourse system, and it is basically measurable. Therefore, the significance of this is even greater.
I believe this is a very important advocacy initiative by the state. As large models become popular, explode in adoption, and then become widely used, it is inevitable that we should propose a unit that can unify the various factor markets and industrial evaluation frameworks. It is a very important economic concept.
【Anchor】Alright. Over the past two years, the average daily usage of token units in mainland China has grown rapidly. With concept mainstreaming and policy support, Mr. Gu, what is your outlook on the future growth scale of the "CiYuan economy"?
【Gu】The thousandfold growth over the past two years has mainly been driven by lightweight interactions such as text-based dialogues and content generation. But now we are seeing that CiYuan usage is surging, which clearly signals that AI applications are evolving from being capable of dialogue to becoming agents capable of making decisions and executing tasks.
An AI agent's CiYuan consumption for a single task is typically tens of times, even hundreds of times, higher than that of dialogue-based applications. For example, an automated research report generation task, a multi-step browser operation, or autonomous code debugging—all of these involve massive CiYuan consumption behind the scenes. This means that over the next two to three years, even if the growth rate of CiYuan usage slows down somewhat, the absolute incremental increase will still be very substantial.
【Anchor】Alright. Then Mr. Jiang, how do you predict the growth potential of the CiYuan economy?
【Jiang】From today's perspective, I remain extremely optimistic about the future of the CiYuan economy.
First, the market scale will grow exponentially. With the overall development of artificial intelligence, it is expected that within the next three to five years, China's token consumption will grow by several hundred times, forming direct economic value at the trillion-level.
Second, from the perspective of value creation across the industry chain, an ecosystem will form: upstream improvements in computing power efficiency, midstream pricing and trading mechanisms, and downstream application scenarios. Token units will become the "hard currency" of the AI era.
Third, the growth momentum is shifting from C-end entertainment toward B-end industrial empowerment. For enterprises seeking to improve production efficiency, high-frequency and complex CiYuan consumption by corporate customers will become the main engine driving growth. The CiYuan economy will also become a core carrier of new quality productive forces.
【Anchor】Alright, then, how will the explosive growth of the CiYuan economy drive changes at the industry logic level? Mr. Jiang.
【Jiang】The explosive rise of the CiYuan economy is reshaping the underlying logic of industrial economics. It represents a shift from a low-traffic mindset in traditional business models toward a "computing power + token units" mindset.
As a result, the entire value-evaluation framework will change fundamentally. Traditional internet traffic is no longer the only yardstick. CiYuan consumption, the quality of generation, and the conversion rate become the core metrics for measuring corporate competitiveness.
On the supply side, the logic will shift from "selling products" to "selling capabilities." Companies will no longer simply sell software; instead, they will output intelligent token units through APIs and charge based on the number of token units. This requires companies to possess extremely strong model-based reasoning capabilities.
At the same time, division of labor across the industrial chain will become more granular. Around the full lifecycle of token units, new niche tracks will emerge—such as prompt engineering, token compression, and vertical token databases—making the ecosystem both richer and more specialized.
【Anchor】Yes. For different groups—companies, entrepreneurs, and consumers—what does the explosive growth of the CiYuan economy mean for each of them? And in this new stage, will "one-person companies" become more and more common? Mr. Jiang.
【Jiang】I think the popularization of the CiYuaneconomy will profoundly change the economic behavior and organizational forms of micro-level actors.
For companies, it means that the intelligent transformation of operating models will be comprehensive. Companies that can use low-cost token units to replace high-cost human labor will gain excess profits. Digital transformation will become a bottom-line requirement for survival.
For entrepreneurs, barriers will decrease, but competitive dimensions will be upgraded. With powerful AI agents, entrepreneurs can leverage massive computing resources at extremely low cost, which will greatly accelerate the spread of "one-person companies."
For consumers, it means access to more personalized and more real-time intelligent services. Consumers may be able to purchase token packages the way they buy utilities like water and electricity. In that case, intelligent assistants could become standard equipment in everyday life.
【Anchor】Alright. Then, with the explosive growth of the CiYuan economy, what are the impacts on enterprises, entrepreneurs, and consumers, respectively? Will "one-person companies" become even more common? Mr. Wang.
【Wang】For consumers, CiYuan will become a pricing factor. Whether you purchase a service or a product, it consumes a certain number of CiYuan. That means CiYuan is a production input. The cost of CiYuan is therefore a very important element for pricing.
In other words, this effectively connects enterprises, entrepreneurs, and consumers in terms of how they measure productivity, prices, and purchasing power.
【Anchor】Yes, just as the industrial era used "kilowatt-hours" to measure electricity, the intelligent era uses "CiYuan" to benchmark intelligence. I'd like to ask Mr. Jiang, at this stage, what do you think are China's competitive advantages? And what uncertainties does it face?
【Jiang】If we measure intelligence with CiYuan, China is currently at a critical turning point from "following" to "leading." Opportunities and challenges coexist.
China's core advantage lies in its massive application scenarios and its extreme cost control capabilities. It has the world's largest CiYuan consumption market. Benefiting from efficient model architectures and energy advantages, domestic models are highly competitive internationally in terms of token-generation cost.
From a policy perspective, the synergy between policy guidance and infrastructure development is also very significant. Current initiatives in China—such as "East Data to West Computing"—provide a solid foundation for the CiYuan economy.
However, challenges are also clear, mainly relating to supply security for foundational computing power and a structural shortage of high-end talent. In a complex geopolitical environment, ensuring a stable supply of high-performance chips is one key question. Another is how to make an effective leap from application innovation to foundational algorithm innovation. These are hurdles that we must cross to develop the intelligent economy and the CiYuan economy.
【Anchor】OK. Regarding "East Data to West Computing," some analysis argue that it can break bottlenecks in computing resources and reshape the industrial chain. However, there are also reports suggesting that some smaller-scale intelligent computing centers operate at less than 30% overall utilization. Why do you think this is the case? And will Western regions gain new development opportunities from East Data to West Computing?
【Jiang】From today's perspective, East Data to West Computing is essentially an optimization and restructuring of production factors across geography. Western regions have access to relatively inexpensive green energy and cooler climates, which gives them a comparative advantage for data centers.
By relocating computing demand in an orderly way to the West, East Data to West Computing can not only reduce society-wide computing costs, but also transform western resources into industrial advantages—creating new economic growth poles. Through this strategy, it can break the constraints of traditional industrial chains and reshape the division of labor in the value chain.
Computing power, as the core productive force of the new data era, when concentrated in western regions, will work like a magnet—attracting upstream and downstream enterprises in data, AI, and related industries to cluster there. This will drive western regions to upgrade from simply being "data storage bases" to becoming centers for data processing and value creation. That is the value-added growth driven by endogenous industrial development.
As for the underutilization of intelligent computing centers, the reasons are more likely structural issues in the short term. For the entire market right now, large AI models have moved from the training stage to the inference stage. Market demand is becoming increasingly fragmented. The intelligent computing centers that have already been built have large reserves of computing capacity. They can fully support lightweight, low-cost, and flexible "organic allocation," enabling effective breakthroughs across the market.
So,in the long run, western computing centers will need to innovate by shifting from traditional resource leasing to scenario-customized models. This would involve diverting training tasks from the east to the west, while simultaneously deploying inference nodes at the edge of cities to meet the business needs of different markets. In doing so, idle computing capacity can be activated,and overall operational efficiency can be improved.
【Anchor】Alright, Mr. Gu, what do you think is China's core advantage in the era of the CiYuan economy?
【Gu】The "three-in-one" advantage of infrastructure is China's most unique structural advantage. The production of token units is, essentially, a transformation process involving three elements: electricity, computing power, and data. Across these three dimensions, China has scale advantages.
China has the world's most complete industrial ecosystem and the largest consumer market. This means that CiYuan application scenarios are naturally abundant. Moreover, China's large AI model CiYuan usage has surpassed that of the United States for three consecutive weeks. The significance of this data is not only in the magnitude of usage itself, but also in the fact that China's demand structure has shifted from "trying things out" to "real demand," and from the C-end to the B-end.
【Anchor】Alright. Then what potential risks do you think there are?
【Gu】Although domestic chips are accelerating their substitution, there is still a gap in the most advanced process nodes and large-scale cluster interconnect—especially in high-performance networking. This directly affects the supply capability for future high-complexity token units, such as multimodal generation and context reasoning.
The thousandfold growth in usage is encouraging, but whether all of it translates into high-quality product value is a question that still needs to be verified over time.
【Anchor】OK, thank you to all the guests. That's all for this episode. Remember to follow us on YouTube or download our APP. Thanks for watching, and see you next time.
Anchor: Laura Cheung | Edited: Kelly Yang, Laura Cheung, Rachel Liu | Translate: Kato Ip | Proofread: Chris Liu
Related News:
DDN Business Insider | Oil hits US$100 as US-Iran talks stall: Experts see long-term highs
Comment