Source: Shenlan Finance

At the recently concluded 2026 World Artificial Intelligence Conference (WAIC), China’s three telecom giants—China Mobile, China Telecom, and China Unicom—took an unprecedented step by converging on a shared narrative: the “Token economy.”

In essence, they aim to position themselves as the “shovel sellers” in the AI gold rush.

This synchronized move underscores the explosive growth of the Token market. By March 2026, China’s average daily Token invocation volume had surged to 140 trillion, representing a staggering 1,000-fold increase since early 2024.

The data signals a clear paradigm shift: Tokens are replacing data traffic as the new currency of the AI industry.

So, who are the “shovel sellers” in this Token-driven economy? And how will Tokens redefine the competitive landscape of AI?

1

The Telecom Titans’ Strategic Gambit

While all three operators are betting on Tokens, their approaches differ.

China Telecom leads the charge. It was the first among the trio to formalize a Token strategy, with Chairman Ke Ruiwen defining Token operations as “delivering AI services to customers” at WAIC 2026.

On the supply side, China Telecom aspires to be more than a reseller—it wants to build “Token factories.” Its Ningxia branch recently launched a centralized procurement project for Token generation capacity, with a projected investment of ¥16.451 billion in 2026 alone.

On the service side, it unveiled the Xingchen TokenHub platform, acting as a “master valve” for Token services. Enterprises can now access diverse AI capabilities through a single Token account and API, eliminating the need for separate integrations with multiple model providers.

China Mobile, meanwhile, centers its Token strategy on industrial internet. During WAIC 2026, it launched “Mobile Tiangong,” a unified industrial internet brand. Leveraging its open platform, it attracts industrial software vendors, smart equipment makers, and system integrators to encapsulate AI models, industry agents, and data tools into standardized, on-demand services billed per Token.

China Unicom adopts a “full-stack Token operations” model, proposing an “Agent + Token + AI Cloud” framework to cover the entire Token lifecycle—from creation and storage to transfer, pricing, and application. It also focuses on cost reduction, developing the Uni-Infer framework to slash inference costs through computing-network-model collaboration, model-computing fusion, and intelligent multi-model selection.

The telecom operators’ collective pivot sends a resounding message: Tokens are no longer a niche tech concept but a core production factor in AI.

2

The Third Wave of Computing Power Monetization

The “Token factory” concept, popularized by NVIDIA CEO Jensen Huang, redefines AI data centers as “modern factories producing tokens.” Each Token, he argues, can be transformed into code, answers, design solutions, or automated decisions—directly translating into commercial value. This vision resonated across WAIC 2026.

Token factories represent the third phase of computing power monetization:

• First wave: Selling hardware (GPU servers).
• Second wave: Selling computing power (cloud providers billed by usage time).
• Third wave: Selling Tokens (customers care only about cost per million Tokens and response latency).

This shift is driven by explosive Token consumption. In January 2025, China’s enterprise-level MaaS market averaged 1.6 trillion daily Token invocations; by December 2025, this soared to 9.6 trillion. For 2026, projections exceed 40,000 trillion—a 20-fold increase.

As AI agents replace conversational AI, Token consumption per task has skyrocketed from thousands to millions, opening vast untapped markets.

This gold rush has attracted not just telecom operators but also cloud giants like SenseTime and hardware leaders like Sugon and Huawei. Meanwhile, independent “Token factories” are emerging with strong capital backing.

On June 30, Silicon Flow filed for a Hong Kong Stock Exchange listing, aiming to become the first Token factory concept stock. Prior to this, it had raised seven rounds of funding, valuing the company at ¥7.74 billion.

Other players are following suit. Tsinghua-affiliated Qujing Technology secured over ¥1 billion in six months, while Infinigence raised ¥700 million in additional funding, bringing its total to ¥2.2 billion.

3

Reshaping the AI Ecosystem

The telecom operators’ synchronized move at WAIC 2026 marks the Token economy’s transition from concept to reality. It also signals China’s AI industry’s maturation: infrastructure is now “energized.”

Previously, firms scrambled for computing power to train proprietary models. Today, with foundational large models stabilized, the focus has shifted to “how to use them”—creating soaring demand for Tokens.

For the industry, Token factories offer structural benefits:

• Upstream: The Token model reduces reliance on NVIDIA’s hardware dominance, enabling domestic chipmakers like Cambricon and Biren to compete on cost and system optimization. System integrators like Inspur Information and Sugon also gain prominence.
• Downstream: Tokens democratize AI, transforming it from a “luxury” to a “utility.” SMEs can now mitigate the risk of “buying the wrong cards” by paying flexible Token fees, unlocking innovation across the ecosystem.

Yet, selling Tokens is no simple task. Operators must balance technical stability with cost control. Mastery of scheduling algorithms and sales forecasting can yield higher margins than traditional hardware sales—but missteps carry greater risks.

4

Conclusion

The shift from “selling computing power” to “selling Tokens” marks the AI industry’s “second half,” as declared at WAIC 2026. The telecom operators’ pivot is just the beginning.

As Tokens become the “digital hard currency” of AI, winners will be those who produce higher-value Tokens at lower costs and greater efficiency. The ultimate prize? AI becoming as ubiquitous and essential as water and electricity.

*Disclaimer: The views expressed herein are solely those of the author. The market carries risks; invest cautiously. This article does not constitute investment advice.

Shenlan Finance, a 16-year-old media outlet, focuses on China’s most valuable companies, cutting-edge industries, and emerging regional economies. Follow us for insights tailored to investors, executives, and the middle class.