Spring Festival AI battle is just a performance? JP Morgan: Token consumption will enter a period of rapid growth, potentially increasing 370 times in five years!

JP Morgan pointed out that although major tech giants have launched fierce AI application promotion battles during the Spring Festival, this is merely a surface phenomenon. Beyond the short-term game of "who will win the AI application war," the market should focus on deeper trends: the way consumers acquire information and consume content is undergoing structural changes, which will drive a multi-year period of rapid growth in Token consumption.

During the Spring Festival of 2026, Tencent, Alibaba, and Baidu will compete for user mindshare through subsidies worth billions of yuan. According to the Chase Wind Trading Desk, on February 6, JP Morgan analyzed in its latest report on the Chinese internet and AI industry that these large internet platforms are investing real money in promotions to accelerate the shift of users from traditional search to chatbots and to cultivate commercial habits around intelligent agents. This process is actually boosting inference volume, thereby accelerating the consumption rate of Tokens. The current situation of excess or insufficient demand for GPU computing power is expected to continue in the coming years.

The bank predicts that from 2025 to 2030, the annual compound growth rate of Token consumption in China will reach around 330%. Based on this, JP Morgan believes it is still too early to judge the winners and losers on the AI application side; more certain investment opportunities lie in the "second-tier winners." This includes direct beneficiaries of China's AI infrastructure (such as Alibaba and Baidu) and companies with exposure to online advertising business (such as Kuaishou). For investors, China's AI infrastructure has become one of the most attractive investment themes in the TMT sector.

Giant's Spring Festival Offensive: Competing for User Habits Rather Than Just Traffic

Major tech companies will adopt vastly different strategies during the Spring Festival, with the core goal of seizing user entry points and mindshare in the AI era.

According to JP Morgan's observations, Tencent has adopted a classic "growth-first" strategy, investing 1 billion yuan in a red envelope campaign. It aims to lower user thresholds by leveraging WeChat's social track, intending to rapidly expand the installation and activation of "Yuanbao."

However, this campaign also exposed the boundaries of platform governance, as some inducements to share led to link restrictions. Analysts believe that Tencent may need to shift from link-driven viral dissemination to product-driven retention loops in the future, integrating AI into the WeChat ecosystem in a smoother, experience-first manner.

Alibaba's strategy, on the other hand, shows structural differences. Alibaba announced a 3 billion yuan "Spring Festival Treat Plan," integrating Taobao, Alipay, Fliggy, and Amap. JP Morgan pointed out that Alibaba's goal is not only to increase the adoption rate of "Tongyi Qianwen" but also to cultivate the commercial habit of "initiating transactions from the AI interface." The strategic aim is to shorten the consumption funnel and improve conversion efficiency; if the effects can be maintained after the holiday, it will help Alibaba occupy an upstream position in high-friction categories such as travel planning. In comparison, Baidu invested approximately 500 million yuan, continuing to adhere to the comprehensive strategy of "AI Assistant + Search Distribution," attempting to embed AI into more intent-driven search conversations. Although ByteDance has not announced large-scale cash subsidies, its Doubao series models reportedly reached 100 million daily active users, thanks to product iterations and the distribution capabilities of ecological entrances like Douyin. JP Morgan believes that even though the monetization model is still uncertain, the cost of losing mental share at the traffic entry level is extremely high, which is a key reason why giants are willing to "buy time" through marketing.

Token Consumption Outlook: A Infrastructure Feast with 370 Times Growth in Five Years

JP Morgan provided astonishing growth forecasts in its report: China's AI inference Token consumption is expected to grow from approximately 10 trillion in 2025 to about 39 trillion in 2030, increasing by about 370 times over five years.

This explosive growth is primarily driven by two factors: first, the increasing penetration of AI in consumer and enterprise workloads; second, the expansion of application scenarios from simple conversational AI to complex agents and multimodal outputs. Data shows that by December 2025, the daily Token consumption of ByteDance's Doubao series models has surpassed 50 trillion, with a year-on-year growth of over 10 times.

Structurally, the composition of inference demand will undergo significant changes. The proportion of conversational AI in total consumption is expected to decline from nearly half in 2025 to a high single digit by 2030. Instead, "knowledge worker AI agents" and "multimodal AI" will take its place. The former, involving multi-step reasoning and document workflows, is expected to account for one-third of total consumption by 2030; the latter, with the increase in image, audio, and video workloads, will grow from the current 15% to nearly 40%.

It is worth noting that JP Morgan reminds investors that Token growth does not linearly equate to chip demand growth. With the inference performance improvements brought by next-generation hardware like Nvidia's Blackwell (e.g., 4 times that of Hopper), and the optimization of the software stack, the amount of Tokens processed per GPU hour is also significantly improving.

Risk Warning and Disclaimer

The market has risks, and investment should be cautious. This article does not constitute personal investment advice and does not take into account the specific investment goals, financial situation, or needs of individual users. Users should consider whether any opinions, views, or conclusions in this article align with their specific circumstances. Investing based on this is at one's own risk