Recently, the National Data Administration named the basic concept Token specification in the field of natural language processing as "lexical elements". Lexicon is the basic unit of artificial intelligence models for processing information such as text, code, and image descriptions, such as a Chinese character, punctuation, or a lexical fragment. With the continuous and rapid growth of the scale of word element calling, a new economic form formed around word element measurement, calling, accounting, and value transformation - the "word element economy" is accelerating its formation. The so-called "lexical economy" refers to a resource allocation method formed around model calling, information processing, cost accounting, service pricing, and value transformation based on lexical units in the operation and application of generative artificial intelligence. At the beginning of 2024, the daily average usage of lexical elements in China was 100 billion, and by March 2026, it had exceeded 14 trillion, an increase of over a thousand times in two years. The key to the accelerated formation of the lexical economy lies in the acceleration of the practical application of generative artificial intelligence. The changes in cost structure on the supply side and the expansion of application scenarios on the demand side provide a realistic foundation for it. Currently, the rapid development of generative artificial intelligence has profoundly changed the way knowledge production, content generation, industrial organization, and value realization are carried out. From the supply side, with the continuous embedding of large model capabilities into specific business scenarios through interface calls and other means, the supply mode of intelligent services is also changing. On demand access, billing according to volume, and dynamic optimization are becoming important trends. From the demand side, intelligent agent applications such as software development, in-depth research, and personal assistants are accelerating their implementation, promoting the extension of artificial intelligence applications from simple question answering to more complex task processing scenarios. This type of scenario usually has longer context, more interaction rounds, and more complex task decomposition. The consumption of lexical elements is significantly higher than that of general interaction scenarios, which also drives the rapid growth of lexical tone usage. Of course, the development of the lexical economy does not solely depend on the expansion of call scale, but more importantly, whether the consumption of unit lexical elements can be effectively converted into real service value and bring efficiency improvement. Essentially, the significance of the lexical economy lies not only in the technical measurement level, but also in the important practical value demonstrated in the value measurement, resource allocation, and large-scale application of artificial intelligence services. Firstly, to make the supply of intelligent services more refined and flexible. Using lexical elements as a measurement scale, relatively abstract model capabilities can be transformed into quantifiable, comparable, and assessable standard service forms. This also makes the cost accounting, service pricing, and resource allocation of intelligent services more refined, and the supply mode more flexible, which can adapt to diverse and differentiated application needs. For users, personalized selection and configuration of intelligent services can be made based on call efficiency, service quality, and cost control. Secondly, it effectively reduces the threshold for accessing intelligent services and broadens the scope of application coverage. Lexicometric enables more small and medium-sized enterprises and developers to flexibly call models according to their actual needs, without having to bear the high cost of whole package procurement, and can achieve on-demand access. This not only helps to promote the extension of intelligent services from a few high threshold scenarios to a wider range of application areas, but also creates realistic conditions for the landing of the keyword economy in more industries and scenarios. Thirdly, promote the collaborative optimization of the intelligent service ecosystem. From a macro perspective, the large-scale application of the lexical economy promotes the construction of high-quality datasets, optimization of computing resources, and iteration of model capabilities, forming a virtuous cycle. The continuous growth of the scale of word element calls has put forward higher requirements for the quality of data supply, which in turn drives the continuous improvement of the data element market and provides a more solid foundation for model training and application implementation. In addition, the refinement of word element measurement has also transformed computing power resources from extensive allocation to on-demand scheduling, thereby improving the overall efficiency of computing power resource utilization. Of course, it should also be noted that the lexical economy, as a new form of intelligent economy that is still accelerating its evolution, still faces some practical challenges that cannot be ignored in its development process. One is the pressure on computing power and energy supply. The large-scale development of the lexical economy means massive model calls and inference calculations, which puts high demands on computing infrastructure and energy security. From a practical perspective, there are still many shortcomings in terms of high-quality data supply, inference computing power guarantee, and open source model ecosystem construction. With the rapid growth of the scale of word element calling, the consumption of computing power and energy demand are synchronously increasing. Whether the supply of underlying resources can continue to form effective support will directly affect the expansion speed, operational efficiency, and cost level of the word element economy. Secondly, the industrial ecology and standard system are not yet mature. The maturity of the lexical economy requires not only demand driven and technological support, but also corresponding industrial support and institutional norms. At present, the supporting infrastructure such as developer ecology and toolchain system is still not sound enough, and the rules around keyword measurement, service pricing, quality evaluation, settlement methods, etc. are still in the exploratory stage. There is a lack of unified and transparent market rules between cross models and cross modalities. The third challenge is facing data security and compliance. The high-frequency calling and continuous interaction of lexical elements are built on the basis of high-frequency and large-scale information processing and data flow, which increases the protection pressure faced by personal privacy, trade secrets, important data, etc. Especially with the increasing number of cross regional and even cross-border calls, the definition of data flow, usage boundaries, and compliance responsibilities has become more complex. It is urgent to accelerate the construction of a sound data security governance framework and cross-border regulatory cooperation mechanism. To promote the healthy and sustainable development of the lexical economy, it is necessary to work together around three aspects. One is to solidify the underlying foundation and break the resource constraints of scale development. Accelerate the construction of computing power infrastructure, especially strengthen the deployment of computing power for large-scale model inference needs, improve the efficiency of computing power supply and resource coordination capabilities, and provide a more solid foundation for the large-scale development of the economy. Continuously improving the supply level of high-quality datasets, supporting the development of open source model ecosystems, and continuously enhancing model training, inference optimization, and intelligent service supply capabilities. We will coordinate and promote the collaborative guarantee of key resources such as computing power and energy, optimize resource allocation methods, enhance the stability and sustainability of the underlying supply system, and better meet the practical needs brought about by the continuous growth of the scale of word element calls. The second is to improve the industrial ecology, perfect the standard system and market rules. Accelerate the establishment of a basic rule system covering vocabulary measurement, service pricing, quality evaluation, settlement methods, and other contents, promote the formation of a unified standard system across models and modalities, make relevant links comparable, connected, and verifiable, and gradually build a unified, transparent, standardized, and orderly market environment. At the same time, it is necessary to better leverage market mechanisms and industry synergies, improve the developer ecosystem and toolchain system, reduce the cost of access and innovation for business entities, and enhance the synergy and resilience of the keyword economy ecosystem. The third is to establish a solid security bottom line and improve the full chain governance mechanism for data security. Accelerate the clarification of data classification and grading protection requirements in high-frequency word element interaction scenarios, build a clear and technically controllable security governance framework, and focus on strengthening the protection of personal privacy and trade secrets. Establish a sound AI generated content review and traceability mechanism, clarify the "measurement tool" attribute of lexical elements, prevent risks such as technological abuse and illegal financial speculation, and create a safe and trustworthy environment for the standardized and orderly development of the lexical element economy. (New Society)