Lenovo: Only tablet maker to reach highest L3 level under China's AI intelligence classification standard

Market Context: Steady Growth with AI as the Core Driver

According to IDC data, global tablet shipments reached 151.9 million units in 2025, representing 5% year-on-year growth. In China, tablet shipments totaled 33.76 million units, up 13.1% year-on-year. Both global and Chinese markets achieved steady growth, signaling the continued expansion of the tablet industry.

IDC further projects that by 2026, global AI tablet penetration will rise to 74.3% — positioning AI as the core variable in the tablet market.

Lenovo's Standout Performance: A Structural Shift in Value

Within this industry trend, Lenovo's tablet business has delivered exceptional results:

• Global market share: 11.8% in Q2
• China market growth: Sales volume up 70% year-on-year; revenue up 165% year-on-year
• Revenue-to-volume ratio: Revenue growth outpaced volume growth by a factor of 2.36

In a market experiencing short-term fluctuations, increasingly rational consumer decision-making, and declining effectiveness of price competition, this divergence is not accidental — it is structural. The average unit value is systematically rising.

The question is: What has Lenovo done right to shift from "winning on volume" to "winning on value"?

Technology Moat: The Only L3-Certified Manufacturer

Following the release of the Ministry of Industry and Information Technology's national standard for AI terminal intelligence grading, Lenovo has become the only tablet manufacturer to achieve the highest L3 level under the national AI standard. This represents a significant barrier to entry.

What makes this threshold so high?

The national standard test includes a task: "Use a shopping app to repurchase movie tickets previously bought." This seemingly simple instruction actually requires a complete orchestration and management chain spanning underlying software, hardware, and applications. Lenovo's device completed the entire process in a fully automated closed loop.

The prerequisite for this capability is a system-level reconfiguration of the three core elements: data, computing power, and algorithms. Lenovo's Tianxi AI enables orchestration and scheduling between on-device and cloud models. The MCP protocol allows seamless invocation of system services, enabling cross-application automation and multi-device interaction — truly achieving "one system, multiple endpoints." This is not a capability that can be acquired through procurement.

Jia Zhaohui, Senior Vice President of Lenovo Group and Head of the Global Innovation Center, stated: "The L3 national standard is not merely a product testing benchmark. It is a yardstick for AI terminal devices to transition from the era of tools to the era of intelligent companions. This is likely a mandatory question for all smart terminal manufacturers on the AI examination path."

The "only" window is also the premium pricing window.

Cost-Side Breakthrough: How "End-Edge-Cloud" Enables AI for All

AI usage cost is a top concern for consumers but is often underestimated by manufacturers. Cloud-based large language model token costs represent a continuous cash outflow. If AI capabilities rely entirely on cloud invocation, the larger the user base and the higher the usage frequency, the less marginal costs converge — this is the root cause of many AI products' unprofitable business models.

Tianxi AI 4.3's "end-edge-cloud" hybrid architecture directly addresses this issue:

• End-edge-cloud collaborative intelligent scheduling: High-level data is stored in private spaces on the device or edge, addressing data security concerns
• Daily tasks handled by tablets: Complex tasks are maximized for local execution on edge devices like Lenovo AI hosts
• Broader applications and services: Handled by the cloud, saving users on token costs

According to Zhang Hua, Senior Vice President of Lenovo Group and General Manager of China Consumer Business Group, Tianxi AI will iterate to version 4.5 in October, with its own programming capabilities integrated.

Jia Zhaohui summarized AI accessibility as three "uses": usable, affordable, and enjoyable. On "affordable," he noted: "Currently, cloud tokens are very expensive. In the future, we need to further develop computing power so that ordinary people can afford it."

Strategic Depth: Not Just Selling a Tablet, but Competing on Ecosystem

Zhang Hua's positioning for Tianxi AI is clear: "Our goal is to make Tianxi AI the total entry point for Lenovo AI device users to connect with AI and devices — the total dispatcher, commander, and coordinator."

This is supported by "three integrations":

• Unified knowledge base invocation
• Shared context across all devices
• Real-time invocation of PC computing power by tablets, with seamless results transmission

The commercial value of this design lies in cross-category synergy. Once users enter the ecosystem, switching costs between PCs, tablets, and phones become retention advantages. Single-product competition is elevated to ecosystem competition, extending customer lifetime value.

The tablet's role within this system has also been redefined. Zhang Hua's characterization: "The AI tablet is a natural container for AI." The goal is equally clear: "If the tablet market were a classroom, Lenovo wants to be the top student."

User-Driven Innovation: The Legion Series as Market Validation

Li Weichang, Vice President of Lenovo Group and General Manager of China Consumer Personal Computing and Smart Terminal Business, stated at the conference: "Competition comes down to two paths — one is low pricing, aggressively capturing the low-end market for rapid growth; the other is doing solid work." Lenovo chose the latter.

He then defined "high-end": "High-end does not mean pricing existing products higher. It means creating additional value for customers that other manufacturers have not solved — value that customers are willing to pay for."

This definition reframes high-end from a "price increase action" to a "value creation action." The former is a short-term margin game; the latter earns genuine customer willingness to pay.

The Legion Y700 Wuji case study:

• Over 1,000 rounds of research during R&D and definition phases
• Product details and customer value were validated item by item
• Decisions on 5G support, screen size, and refresh rate were crowdsourced from 100,000 fans in the Super Core Community
• Lenovo leveraged industry resources to bundle 1,500 GB of free monthly data

The result: The Legion Y700 became a phenomenon — sold out within 15 minutes of launch, 100% positive reviews, and ranked No.1 across all platforms.

The significance of this process: Demand certainty was front-loaded to the R&D stage. Traditional approaches involve project initiation, production, and then market testing — with inventory risk and misjudgment costs borne at the back end. By incorporating user voting into the definition phase, Lenovo effectively eliminated demand-side uncertainty through community engagement.

This improves two operational metrics simultaneously: new product success rate and inventory turnover.

The Y900 series, also built on user insights, received positive market feedback: over 50% of initial buyers were college students, and back-to-school season sales increased 30.8%. The user-driven innovation closed loop is being validated by product data.

Li Weichang also revealed during a group interview that from September 23 to October 7, all students purchasing Lenovo AI tablets — especially the Legion tablet — would receive exclusive student discounts. Buyers of the Legion AI tablet would also have a chance to receive a free Legion AI notebook.

Organizational Efficiency: R&D Speed as a Financial Variable

"Traditional R&D, production, supply, sales, and service — the further you go, the farther you are from the customer. Our innovation model is horizontal: R&D, production, supply, sales, and service all face the user horizontally," said Zhang Hua.

Supporting mechanisms include:

• Super Core Community: 100,000 users providing real-time feedback
• Monthly Updates: New experiences pushed every month, with feedback determining which features are high-frequency and which receive negative reviews

The financial reflection of this approach: Lower R&D trial-and-error costs and higher new product hit rates. In an era where technology roadmaps have not yet converged, iteration speed itself is an asset.