Tokens are emerging as a pivotal breakthrough for telecom operators, aiming to reduce their dependence on data traffic revenue.

Amid surging demand for AI inference and rapid Token consumption, the three major telecom operators in China have nearly simultaneously introduced Token packages for individual and enterprise customers. Their goal is to make computing power as easily accessible as mobile data plans—ready to use right out of the box.

China Telecom has taken the lead, being the most proactive and representative among the three. Its most affordable Token package for individual and family customers is priced at 9.9 yuan per month, offering 10 million Tokens (sufficient to generate approximately 2,000 to 5,000 ten-page research reports or 30,000 to 50,000 short video scripts). For developers and enterprise users, the lowest package is 39.9 yuan per month, providing 15 million Tokens.

Unlike China Telecom's unified pricing strategy, China Mobile and China Unicom are adopting a province-by-province approach and have yet to establish unified computing power packages at the group level. For instance, several provincial branches of China Mobile, including those in Guangdong, Jiangsu, and Beijing, began trialing Token packages in April. Similarly, China Unicom initially rolled out Token packages through its Shanghai and Hubei branches.

The specific outcomes of the operators' initiatives remain to be seen, but they have already sparked significant activity in the capital markets. China Telecom's stock price has surged for two consecutive days since the Token package announcement, while China Mobile and China Unicom have also experienced notable gains.

01

Is 9.9 Yuan for 10 Million Tokens Too Pricey?

The launch of Token packages by telecom operators has drawn considerable attention from the media and social networks. However, it appears that the operators may not be fully prepared.

For example, while China Mobile's app allows direct purchase of Token packages, could not find an experience entry after subscribing. When contacting the 10086 customer service hotline for assistance, the representative remained silent for an extended period and was unaware of such products.

A China Mobile employee informed that the group had issued a notice encouraging employees to experience the service and that the customer service department was developing service process standards related to computing power. However, specific scripts had not yet reached frontline staff. The group had also issued guidance but had not yet established concrete performance metrics.

The employee revealed that operators have been exploring a transition to a higher-value Token economy. Earlier, they planned to launch a "crayfish package," which shifted to selling Tokens within two months and now combines Tokens with services. "Things keep changing, and we're adjusting too," he said. Currently, not only users but also many frontline service personnel lack understanding, resulting in high explanation costs during promotion.

However, among the numerous discussions about operators transitioning to selling Tokens, pricing is a major point of user criticism.

Many users note that operators' Token prices are higher than those of many model providers' API calls. For example, Beijing Mobile previously offered a Token package priced at 24.99 yuan per month, including a quota of 10 million Tokens. Shanghai Telecom's Token package, which allows access to over 30 mainstream large models, costs 1 yuan for 250,000 Tokens, with the price per million Tokens reaching 4 yuan.

Moreover, operators' provincial branches have significant pricing authority, leading to varied pricing strategies for Token packages across regions, causing confusion among consumers.

Operators' Token prices are inevitably compared to those of cloud and model providers. Taking the aggressively priced DeepSeek as an example, DeepSeek-V4-Flash costs 0.02 yuan per million Tokens for input cache hits, while DeepSeek-V4-Pro costs 0.025 yuan. For cache misses, the price is 3 yuan per million Tokens.

However, operators do not distinguish between input and output Tokens or differentiate pricing based on cache hits—a common pricing approach in the industry for Token plans.

In fact, on paper, operators' Token package prices are not significantly different from those offered by internet cloud providers. For example, China Telecom's lowest monthly fee is 9.9 yuan (10 million Tokens), while Tencent Cloud's lowest monthly fee is 39 yuan, including 35 million Tokens. However, cloud and model providers generally charge based on the number of requests rather than Tokens, indirectly complicating comparisons.

It is worth noting that current operator Token packages for consumers primarily target "crayfish farming" needs, as evidenced by their usage methods. For instance, China Telecom offers multiple access methods: customers with existing local hardware terminals (e.g., Mac Mini) can set their API keys for Token consumption across various local "crayfish" applications, while those without local hardware can bind directly through cloud computers or cloud phones.

Based on package offerings in some provinces, intelligent agents like "crayfish" that require extensive workflow invocation can consume tens of millions of Tokens daily under heavy usage, representing a significant expense for users.

"We're still adjusting because pricing varies across channels for acquiring computing power. A unified pricing plan will follow, similar to how data plans were initially high and gradually adjusted downward," a China Mobile representative told . Token packages in some regions have been implemented for over a month, and internal evaluations confirm that pricing is indeed high.

The representative revealed that operators will also refine Token packages, such as introducing specialized model Token bundles and general-purpose model Token bundles. They are even exploring industry-specific Token packages for finance, healthcare, and education.

02

What Are the Operators' Strategies?

As national infrastructure builders, telecom operators have faced significant business transformation pressures in recent years. On one hand, user growth has plateaued, and revenue growth has stalled; on the other hand, user ARPU (average revenue per user) has declined annually. For example, China Mobile's mobile ARPU was 46.8 yuan in 2025, down from 49.3 yuan in 2023. China Telecom saw a similar decline, from 46.2 yuan to 45.1 yuan.

Meanwhile, according to data from the National Data Bureau, as of March 2026, China's daily token invocation volume exceeded 140 trillion, growing over 1,000-fold from early 2024, indicating accelerating AI application demand.

For operators seeking transformation, shifting from traditional pipeline services like communications and data to intelligent services like Tokens and computing power is no secret. Tokens are emerging as a core breakthrough for operators to reduce reliance on data traffic. In 2025, China Mobile's computing power service revenue grew by 11.1% year-on-year, with communications services, computing power services, and intelligent services clearly defined as the company's three growth drivers.

China Telecom has also stated that "Token services will be its future business focus," offering differentiated services and pricing based on general-purpose, professional, and customized Tokens.

An industry insider told that a more direct motivation for operators to launch Token packages is their structural computing power redundancy. Low-end general-purpose computing power remains underutilized, and domestic general-purpose cards are difficult to digest. The primary goal of Token packages is to activate existing resources and monetize them.

However, the insider noted that operators' overall computing power construction has not slowed; they are simply shifting from blind expansion to precise deployment of high-end intelligent computing while strictly controlling inefficient capacity.

The biggest challenge for operators lies in their weak model capabilities. Selling Tokens is not entirely analogous to operators selling data traffic in the past. Traffic and networks are franchised operations where operators invest in building base stations, which other enterprises cannot do. However, Tokens essentially represent computing power + models, and nearly all cloud providers offer MaaS (Model as a Service) with corresponding Token package plans.

This means Tokens will not only be a competition among major operators but also face competition from cloud providers, model providers, and other open-market players.

Indeed, the three major operators have launched self-developed foundational models, such as China Mobile's Jiutian model, China Telecom's Xingchen model, and China Unicom's Yuanjing model. However, operators' large models generally cannot directly compete with market models and instead emphasize domestic computing power support to meet the needs of some government and enterprise clients.

From current experiences, even operators' Token packages do not invoke their proprietary models. China Mobile uses Minimax, while China Telecom uses Zhipu GLM and DeepSeek—and not the latest versions, such as DeepSeek V3.2.

In fact, operators have been strengthening their talent pools and betting on new businesses. China Telecom, often seen as a steady yet bold operator, has been breaking state-owned enterprise salary mechanisms to hire top talent in recent years.

Moreover, operators do have their advantages compared to cloud providers.

Operators have built extensive computing power networks and can secure the most favorable electricity resources locally, enabling them to reduce Token costs through economies of scale. China Galaxy Securities points out that operators, with their nationwide backbone network resources, rich computing power node layouts, and "cloud-network integration" capabilities, play a key role in connecting dispersed computing power and breaking geographical barriers.

Telecom industry analyst Yang Guang believes that some price-sensitive but latency-tolerant small and medium-sized enterprises may find Token packages attractive. Additionally, operators are building their model marketplaces, which can allocate different models based on task urgency through model routing. For example, latency-insensitive tasks can be computed in western data centers.

Second, operators have constructed numerous data centers with domestic computing power, aligning with their narrative of "domestic chips, domestic clouds, and domestic models" and tying Token operations to network security, computing security, and data security. While C-end consumers may not feel this strongly, there is strong demand in the government and enterprise market, Yang Guang told .

In fact, operators are gradually shifting from a pure Token-selling model to Token + services, which includes bundling network billing and security services with Tokens.

Furthermore, operators have a nationwide business network and a vast service system that penetrates deep into local markets, giving them unparalleled business touchpoints compared to other vendors.

"Account managers interact with these organizations daily and may identify clients needing such packages," an operator representative said, citing a potential scenario where account managers selling broadband to merchants could also offer a Token package for the client's store cameras.

Operators also have a natural advantage: seamless integration between mobile numbers and Token users, eliminating the need for separate registration and enabling direct Token purchases using phone credits, greatly facilitating Token consumption.

"There will be short-term stimulation, as some users are always willing to try new things," Yang Guang said. However, whether operators can build momentum with Tokens remains to be seen. "Willingness to try is still a good sign."