WPS boasts over 80 million domestic AI monthly active users, but nine out of ten AI practitioners don't know it — Kingsoft Office's buzz severely mismatches its product strength. This week we noticed an update to WPS Multidimensional Table: it supports writing Python directly in cells, opens the MCP protocol, and offers over 70 APIs. This looks less like a feature iteration and more like a strategic pivot.
What this is
WPS Multidimensional Table (a structured data management tool similar to Airtable) showcased several new capabilities at a closed-door sharing session this week:
Python + AirScript dual-script engine: Write Python code directly in the browser, with results rendered in real-time to the table. No need to install environments or configure dependencies. AI can even help write scripts. MCP protocol support (an open protocol allowing AI Agents to call external tools): You can invoke external Agent platforms like Claude Code within WPS Multidimensional Table, with separately controllable permissions. Performance data: Under millions of rows of data and thousand-level concurrency, the average editing response is 32 milliseconds. It ranks second globally in the SpreadsheetBench evaluation, right behind Google Gemini.
The exact words of product lead Yang Ding are noteworthy: "We don't restrict people to necessarily use our stuff; the ultimate goal is to solve users' actual problems."
Industry view
The path WPS is taking differs from Feishu Multidimensional Table. Feishu focuses on competing on AI capabilities themselves — smart autofill, formula recommendations, natural language table creation. WPS competes on openness — letting developers build applications on top of the spreadsheet.
This choice makes sense. When all multidimensional tables can use AI to build tables, the functional gap narrows. The moat is no longer "whose AI is smarter," but "whose platform can grow more things."
But the risks are equally obvious. WPS's buzz in the developer community is almost zero. One attendee bluntly said: "Your voice in the AI circle is basically zero." We asked around our AI friends, and nine out of ten didn't even know WPS had a multidimensional table. A developer ecosystem isn't built just by opening APIs — Feishu Multidimensional Table already has a first-mover advantage, and Notion and Airtable are deeply rooted in overseas markets. For WPS to get developers to migrate from these platforms, technical specs alone won't be enough.
Impact on regular people
For enterprise IT: Traditional manufacturing cases show that a single multidimensional table can replace three systems: OMS + APS + MES. This is a tangible cost reduction for SMEs without IT teams.
For individual careers: Non-technical roles like operations and product management, which previously had to find programmers to run data processing scripts, can now do it themselves within the spreadsheet. The barrier is lowered, but basic Python knowledge remains a plus.
For the consumer market: 80 million AI MAUs show WPS has a massive user base, but the audience for multidimensional tables still skews enterprise. In the short term, it won't become a mass-market product like WPS Docs.