CGS released a research report stating that tokens have become the value anchor in the intelligent era, widely used as a settlement unit in the AI industry. From the demand side, the global token call volume is growing rapidly, driving the AI industry from model iteration to commercialization. This is expected to fundamentally reshape the production factors, industrial logic, and business models of the media industry. The token economy empowers traditional content production industries, and with the support of AIGC tools, it is moving toward unlimited supply, potentially leading to prosperity in the content industry. It is recommended to focus on core AI industry beneficiaries, with key attention on: 1) Internet giants intensifying their comprehensive AI布局; 2) Manufacturers leading in AI video tool product capabilities; 3) Leading companies in细分领域 benefiting from AI applications; 4) Large model manufacturers continuously advancing technological iterations. The main views of CGS are as follows: Tokens are the value anchor in the intelligent era. In March 2026, the China Development Forum officially designated the Chinese name for Token as "词元" (token), positioning it as the value anchor in the intelligent era and a unified settlement unit for both supply and demand sides of the industry. Currently, the daily average token call volume in China is showing an exponential upward trend, marking that China's artificial intelligence industry has moved beyond the model technology verification phase and fully entered the cycle of large-scale computing power consumption and commercial operation. AIGC drives content industrialization, with unlimited supply reshaping the industry landscape. As token consumption continues to grow rapidly, AI推理 demand is fully unleashed, and the media content industry is accelerating its transition from traditional manual creation models to AI-native generation models. Content production is彻底摆脱人力产能的约束,创作效率的瓶颈, and high marginal cost limitations,正式迈入规模化,标准化, and infinitely replicable content industrialization mass production stage. Large models continue technological iteration, with open-source ecosystems driving生态发展. The continuous increase in token call volume directly drives爆发 in API接口调用 demand across various downstream applications, which will further传导至模型底座 manufacturers. The持续放量 of token demand can放大模型底座的复用价值 and有效摊薄 model training and technical运维的单位成本, continuously strengthening the technological壁垒 and market competitive advantages of model底座 manufacturers, becoming the core driver supporting the long-term growth of the model底座 industry.云服务量价共振,有望打开增长新空间. The continuous increase in token consumption corresponds to the large-scale放量 of AI推理 demand across society, constituting确定性利好 for cloud service providers represented by阿里云. As tokens become the unified value settlement unit in the intelligent era, the云计算商业模式 is undergoing fundamental重构. The industry is gradually脱离传统服务器租赁,时长计费的IaaS模式, fully transitioning to a MaaS智能服务按量计费体系 centered on token consumption. This promotes the expansion of high-margin AI service revenue for云厂商,优化整体盈利结构, and the industry enters an上行周期 of demand放量 and量价共振. Risk warnings include risks related to policies and regulatory environments, risks of AI technology development falling short of expectations, risks of AI application落地不及预期, and risks of short-term adjustments in market sentiment and capital flows.
Comments