In recent years, the mobile internet’s rapid evolution has driven telecom operators toward commoditization—a trend now accelerating with the rise of AI. Against this backdrop, China’s three major telecom operators—China Mobile, China Unicom, and China Telecom—have collectively launched token packages. Could selling AI tokens become their next major revenue stream?

I. Operators Unveil Unified Token Packages

According to Xinhua News Agency, AI technology’s rapid iteration and the proliferation of intelligent agents and large-scale model applications have fueled explosive growth in token usage across China. The three operators have rolled out token packages for individual consumers and enterprise clients, aiming to democratize AI access while fostering deeper integration between telecommunications, computing power, and AI industries. This strategic move is creating new opportunities for digital economy development.

Data from the National Data Bureau reveals staggering growth: China’s daily token consumption surged from 100 billion in early 2024 to 100 trillion by late 2025, exceeding 140 trillion by March 2026—a thousandfold increase in two years. As demand for AI computing power soars, operators are leveraging their computing networks, cloud infrastructure, and massive user bases to accelerate token product commercialization.

On May 17, China Telecom launched trial commercial token packages, offering multiple tiers for individuals and developers, with plans starting at 9.9 yuan/month for 10 million tokens. On May 16, China Unicom Shanghai Branch provided local OPC (One-Person Company) users with 30 million free test tokens. In April, China Mobile introduced token packages in Beijing, Hubei, Henan, and other regions...

“A few interactions with an AI agent can burn through tens of thousands of tokens,” said Beijing resident Mr. Su. Many users echo his experience, citing high computing costs, cumbersome model access, and fragmented payment systems when using large-scale models, AI creation tools, and intelligent agents.

II. Can Token Sales Transform Operator Profitability?

China’s telecom giants have strategically prioritized “token operations,” launching tariff packages for individuals and enterprises. What drives this coordinated shift? Can it truly redefine their revenue models?

1. Escaping Commoditization: A Necessary Evolution
For decades, operators’ profits relied on voice, SMS, and data traffic. During the 2G era, packages like “30MB for 5 yuan” were key revenue drivers. However, the 4G/5G era has eroded this model: data traffic prices have plummeted under policies mandating “faster speeds, lower fees,” turning it from a scarce resource into a commodity. Per-user data usage growth has plateaued, while revenue per user declines annually. Meanwhile, internet platforms dominate the value chain, leaving operators as mere “data conduits” earning meager channel fees.

This commoditization crisis reflects the mobile internet’s maturity. Operators’ network infrastructure underpins the digital economy but remains at the bottom of the value distribution chain. Any opportunity to escape this role and rebuild profitability is now critical. The AI boom offers precisely that chance. Tokens, as the currency of AI computing power, hold far greater value than traditional data traffic. Selling tokens represents a leap from “connectivity” to “intelligence”—a strategic bid to redefine industry positioning.

2. AI Agents Drive Token Demand
AI intelligent agents have exploded in popularity this year, with trends like “virtual pet raising” going mainstream across industries. Computing power is no longer a backend necessity but a frontend consumption driver. Tokens, the basic unit for large-scale model processing, are seeing exponential consumption growth.

This shift offers operators hope: Just as users paid for “bandwidth” in the mobile internet era, they now have reason to pay for “computing power” in the AI era. Operators recognize tokens as the “new traffic” of AI—standardized, measurable, and高频交易 (high-frequency tradable). By packaging AI capabilities into prepaid-like plans, they’ve democratized cutting-edge tech, making it accessible to consumers and SMEs. This opens doors to a trillion-yuan computing power market.

3. Unmatched Operational Advantages
The three operators boast irreplaceable strengths in this transformation: nationwide computing networks and unparalleled channel reach. They own backbone fiber networks, data centers, and edge computing nodes across China, ensuring low-latency AI inference. According to China Business Journal, AI-related computing investments and revenues dominated operators’ 2025 highlights: China Telecom invested 20.2 billion yuan (25% of total) in computing infrastructure; China Mobile’s computing service revenue hit 89.8 billion yuan (up 11.1% YoY, 20.2% of total); China Unicom’s computing business exceeded 15% of revenue.

Critically, operators control 1.213 billion 5G users (as of April 2026) and millions of enterprise clients. For most users, AI adoption barriers include complex registration, foreign payments, or technical jargon. Operator token packages integrate with prepaid systems, letting users buy AI computing power as easily as topping up phone credit. This convenience and trust lower AI adoption thresholds, enabling operators to capture lower-tier markets and long-tail users—a moat internet giants can’t replicate through short-term subsidies.

4. Challenges Ahead
Despite operators’ strengths, token business risks loom. Whether token sales can drive sustainable profits remains uncertain. Tokens, tied to computing power and energy, face constant depreciation. Their marginal costs—GPU computing and electricity—rise with usage, pressuring operators to control expenses. A misstep could lead to “revenue growth but profit decline.”

Additionally, token value varies by model. The same 1 million tokens yield vastly different intelligence levels on small vs. top-tier models, affecting market pricing. Operators currently act as “token resellers,” reliant on external core models. Without mastering core R&D, they risk becoming low-margin channel partners for model providers.

While token sales appear lucrative, their logic differs fundamentally from mobile traffic. Whether this can fuel new growth is unclear. The token economy holds promise, but profitability isn’t guaranteed. For operators, breaking revenue stagnation with tokens will test their strategic agility.

END