从DOS时代到AI办公时代 WPS凭什么陪了一代又一代人？

In the age of AI-powered office work, WPS has launched Lingxi Professional Edition for individuals and WPS Comate for organizations. With over 676 million global monthly active users and more than 80 million active AI users, it has surpassed Microsoft in office suite market share. Here are the key points for general readers:

1. AI that gets work done is what truly matters. Being able to chat and being able to work are two different things. A useful office AI must understand a company's own processes, data, rules, and boundaries, not just write documents, make spreadsheets, or generate PPTs.

2. Personal productivity starts with your own documents. When a company’s documents, contracts, reports, emails, and meeting minutes are combined, they form a record of how the company operates and the foundation for AI to understand the business. The more documents accumulated in WPS 365, the more accurate the AI's assistance becomes.

3. Choose the right scenarios for AI. Pick scenarios that are high-frequency, data-dense, and involve many people, money, and time. Don’t expect AI to solve everything; AI projects cannot be left unattended—they require close collaboration with people who understand the business.

4. Data security can be trusted. WPS Comate is built on security and permission rules at its core, and has been awarded the first national-level intelligent agent identity code node qualification in the AI office industry.

WPS has evolved from the DOS era to the AI office era. In the first quarter of 2026, it surpassed Microsoft with a 32.1% market share versus Microsoft's 28.5%. Its brand building, product R&D, and competitive strategy offer lessons for brand owners.

1. Brands grow by accompanying users. WPS has served the entire informatization journey of Chinese enterprises, witnessing the evolution from paper-based office work to standalone computers, networked systems, subscription models, digitalization, and now intelligentization. Having seen all types and stages of enterprises, this deep accumulation builds brand trust.

2. Product R&D should be phased. Kingsoft Office has developed AI in three generations: AI 1.0 is office component AI, letting employees use AI within familiar tools; 2.0 is knowledge base AI, building enterprise knowledge assets that AI can understand, execute, and govern; 3.0 is AI Office, intelligent agent AI Copilot, and Lingxi, creating an organizational human-machine collaboration platform. The logic is clear: from tools to knowledge to organization.

3. Competition relies on localized context. Large models don’t understand enterprises, but WPS has spent 38 years learning what enterprise knowledge looks like, how data flows, and where rules are hidden. Enterprises choose AI not for a chat window, but for a mechanism that can integrate model capabilities into business processes.

4. User behavior shows that AI office consumption is shifting from individual efficiency to organizational efficiency. Enterprises want AI to become digital employees, not just document tools.

The AI office market is booming. WPS has over 676 million global monthly active users, more than 80 million active AI users, and WPS 365 revenue grew 61% year-over-year. In office suite market share, Kingsoft leads with 32.1%. Sellers should watch these opportunities and risks.

1. Growing market: ByteDance Doubao Work, Tencent WorkBuddy, and Alibaba Qianwen Office have all launched integrated releases. Players in the AI office space are surging, and enterprise-level AI demand is strong.

2. Changing demand: Enterprises are no longer satisfied with AI that can chat; they truly need AI that understands company processes, data, and rules and can get work done directly. Whoever solves context engineering will capture customers.

3. Risk alert: A common dilemma is that enterprise AI investment has not translated into organizational efficiency. The three major bottlenecks are data silos, experience blockages, and action blockages. Selling tools without helping to connect processes makes success difficult.

4. Partnership approach: Within a week of WPS Comate's launch, hundreds of large and medium-sized enterprises expressed co-creation intent, spanning transportation, finance, manufacturing, healthcare, and energy. Kingsoft's implementation experience is to connect data, knowledge, and capabilities; manage security and costs; and unify into an office AI platform. Customers need to work with the FDE team from the outset.

5. Practical points: When selecting scenarios, look for high-frequency, data-dense, and involving many people, money, and time. Feed internal data to the agent so AI output moves from being correct to being usable.

Manufacturing enterprises are being reshaped by AI-powered office tools. WPS has served scenarios such as interconnecting process standards across three factory sites for manufacturing companies. CIMC Group, Kaiwo Automobile, and other manufacturers have become among the first co-creation enterprises of WPS Comate. Factories can draw inspiration from the following.

1. Production data must first become context that AI can understand. When contracts, reports, process standards, approval records, and other daily data are stacked together, they form a record of how a factory makes decisions and operates. To advance digitalization, factories should first accumulate and manage these knowledge assets.

2. The six elements of context engineering are key to AI implementation. Knowledge, data, time, relationships, rules, and boundaries determine whether AI can understand the real business of a factory. For example, the interoperability of process standards, approval rules, and the boundaries of each role must be fed to AI in a structured way.

3. Commercial opportunities are clear. AI office is entering the stage of organization-level human-machine collaboration. WPS Comate connects an organization's data, systems, processes, knowledge, and capability models, moving from providing suggestions to directly executing. Manufacturing enterprises can use such platforms to break down data silos between plants and departments.

4. Implementation insights. Connect data, knowledge, and capabilities; manage security and costs; and unify into one office AI platform. Choose scenarios that are high-frequency, data-dense, and involve many people, money, and time. AI projects cannot be left unattended—work with the implementation team from the start.

Enterprise AI implementation is moving from tools to organizations. WPS Comate provides service providers with a complete sample for understanding customer pain points and solutions. The following points deserve attention.

1. Customer pain points: Large models understand the world but not the company. They don't know the meaning of costs in different departments, approval rules, or procurement bottlenecks. Enterprises' AI investments generally have not translated into organizational efficiency. The three major bottlenecks are data silos, experience blockages, and action blockages.

2. Solution: The core is context engineering, which includes six elements: knowledge, data, time, relationships, rules, and boundaries. Feed professional data, internal information, and simulation data to the agent, and the customer research report will move from looking correct to actually being usable.

3. Architecture reference: WPS Comate is built on security and permission rules at the bottom. Above that are three hubs—API Hub, Data Hub, and AI Hub—then the core capability layer (management experts, Skills, knowledge graphs, team spaces). The top layer supports local and cloud tasks and multi-device access.

4. Implementation experience: Connect data, knowledge, and capabilities; manage security and costs; and unify into one office AI platform. When selecting scenarios, look for high-frequency, data-dense, and involving many people, money, and time. Customers of AI projects cannot be absentee landlords—they need to work with the FDE team from the start, otherwise results will be hard to achieve.

The trend toward platformization in AI office is accelerating. WPS has used 38 years of accumulation to embed AI into an organization-level human-machine collaboration platform, offering a new model for platform operators.

1. Platform demand: Enterprises do not choose an AI platform for a chat window, but for a mechanism that can integrate model capabilities into business processes. The platform must connect an organization's data, systems, processes, knowledge, and capability models, evolving from advisor to executor.

2. Platform architecture: WPS Comate is built on security and permission rules at the bottom. Above that is API Hub to connect business systems, Data Hub for data governance, and AI Hub to manage large model consumption and scheduling. Then comes the core capability layer, and the top layer supports local and cloud tasks and multi-device access.

3. Operations and management: The platform needs to do five things well: manage knowledge and data, connect internal enterprise systems, turn knowledge into experience, manage large model usage and the full lifecycle of experts, and ensure information security.

4. Avoiding pitfalls: Avoid the situation where AI can chat but cannot work; avoid becoming an isolated plugin or a replacement for business systems. The failure of enterprise AI investment to translate into organizational efficiency is a common risk. Platforms should help enterprises connect data, knowledge, and capabilities, and manage security and costs.

5. Latest moves: WPS Lingxi and Comate have received the first national-level intelligent agent identity code node qualification in the AI office industry, becoming the first to obtain the digital identity for agent interconnectivity and collaboration. This is an important exploration of platform compliance and interoperability.

Kingsoft Office's 38-year journey from DOS to AI, along with the release of WPS Comate, provides an important case for studying the AI office industry. Core observations are as follows.

1. New industry trends: The AI office track has many players, but being able to chat and being able to work are two different things. The real bottleneck in enterprise AI implementation is not model capability but context engineering. Large models provide intelligence, while software provides contextual judgment—this distinguishes the critical point between AI products and enterprise applications.

2. New problems: The vast majority of enterprises' AI investments have not translated into organizational efficiency; individual employee productivity does not equal organizational productivity. Data silos, experience blockages, and action blockages are the three major bottlenecks in enterprise AI implementation, offering a problem checklist for future technical and management research.

3. Business model: Kingsoft Office has developed AI in three generations, from office component AI to knowledge base AI, then to intelligent agent AI Copilot and Lingxi. From tools to knowledge to organization, it ultimately moves toward a one-stop AI-native office platform. Agents run on WPS context; the more knowledge assets, the more accurate the context, the more efficient agent execution, forming a growth flywheel.

4. Governance exploration: WPS has received the first national-level intelligent agent identity code node qualification in the AI office industry, securing the digital identity for agent interconnectivity and collaboration. This suggests that intelligent agent identity authentication and interconnected collaboration governance are becoming directions for institutional innovation.

5. Competitive insights: Kingsoft Office's 32.1% market share surpassing Microsoft demonstrates that localized context accumulation and long-term experience in serving Chinese enterprises' informatization are key variables in the competition of Chinese office software.