On March 25, according to press reports, yesterday, at the 2026 annual meeting of the China Development High-Level Forum, the Director General of the National Data Agency, Liu Liang, stated that Token "CharacterIt's not like it's the same thing. Not only are the value anchors of the intellectual age, but it is also a “settlement unit” that connects the supply of technology with commercial demand, providing quantifiable possibilities for the landing of business models。
This is an official translation of Token's Chinese translation: The Word。
1AI UNDERSTANDS THAT IN ARTIFICIAL INTELLIGENCE SYSTEMS, WORDS ARE THE SMALLEST UNIT FOR MODELING INFORMATION. WHETHER A USER ENTERS A QUERY OR A SYSTEM-GENERATED PROGRAM CODE, YOU HAVE TO DISASSEMBLY INTO A WORD TO RUN. AS A RESULT, THE HYPE CALL VOLUME IS AN IMPORTANT INDICATOR NOT ONLY OF THE OPERATIONAL ACTIVITY OF THE MODEL BUT ALSO OF THE REAL VALUE CREATION CAPACITY OF THE ARTIFICIAL INTELLIGENCE INDUSTRY。
Liu Lihhong revealed that, as of March 2026, our daily word calls had risen to more than 140 trillion times, a thousand times higher than the 100 billion at the beginning of 2024; even when compared to 10 trillion at the end of 2025, the increase exceeded 40 per cent in just three months。
Liu Liang noted that since the end of January this year, some model firms have established a record of 20 days of income over and above total annual income in 2025. Behind this set is a new business logic based on Token billing。
In this assessment, Liu Liang stated that the sharp increase in the number of calls reflected the fact that the country ' s artificial intelligence industry was on a high-speed development path. As the application landscape evolves from interactive dialogue to an implementation-of-decision smart body, the competitiveness of China’s artificially intelligent industry is increasing significantly, and the high profile of Token going out to sea is a sign of increased industrial competitiveness。
AT THE LEVEL OF DATA RESOURCES, WE HAVE ACHIEVED MILESTONES IN THE CONSTRUCTION OF HIGH-QUALITY DATA SETS. BY THE END OF 2025, MORE THAN 100,000 HIGH-QUALITY DATA SETS HAD BEEN COMPLETED THROUGHOUT THE COUNTRY, AND THE SIZE OF STORAGE HAD EXCEEDED 890 PBS, EQUIVALENT TO 310 TIMES THE TOTAL DIGITAL RESOURCES OF THE CHINESE NATIONAL LIBRARY。