Alibaba's Qianwen Office Hits 30M Users in First Month; Enterprise Users Top Half

Qianwen Office opened to users one month earlier via dedicated PC clients, a web portal and DingTalk, Alibaba's workplace messaging platform. The product was built by integrating and converting capabilities and customer bases from several earlier Alibaba offerings, including QoderWork, Wukong and MuleRun. From the first day, it was connected to DingTalk across document, spreadsheet, calendar, message and email scenarios.

Industry insiders say reaching 30 million users in such a short period was faster than expected and unlikely to come purely from organic growth. Alibaba's enterprise strengths, especially DingTalk, helped Qianwen Office embed AI into daily work rather than sit as a separate app. An industry observer cited by Leiphone said that, unlike in the internet era, enterprises have to embed AI tools into existing workflows if they want real efficiency gains. Alibaba Cloud's enterprise service capabilities and customer relationships also gave Qianwen Office an advantage: large corporate clients facing AI-transformation pressure were willing to test the product, and successful use cases often spread within the same organization and across its supply chain.

In the past month, Qianwen Office shipped 120 version updates and added more than one thousand features, while rolling out an international edition. It also introduced an organizational skill library, allowing a skill created by one employee to be reused across the entire company, and launched a multi-person workbench that can generate a collaborative online webpage from a single sentence, enabling up to 100 people to work simultaneously.

Qianwen Office is also experimenting with breaking down AI capabilities into smaller, reusable atomic abilities that can be inserted into existing enterprise software. For example, sales personnel no longer need to leave DingTalk to open a separate AI product; they can ask the agent to query customer, inventory and order records across CRM and ERP systems and receive synthesized business recommendations. From the employee's perspective, this adds natural-language interaction; from the company's view, AI starts connecting business systems that have long been siloed.

To deal with what consultants call data rot, where agents drift out of date because underlying information changes quickly, Qianwen Office open-sourced MyContext, a context infrastructure project. MyContext helps agents continuously sense a user's identity, department, project background, customers and related materials instead of forcing manual input every time. Recruitment is an early working scenario: recruiting agents perform resume matching and evaluation inside the original workflow, with human-resources staff involved only in interviews and final decisions.

The product's user structure is seen as its biggest commercial signal. More than half of the 30 million users are enterprises, meaning the product has reached customers with stronger payment capabilities. A market-research report by Frost & Sullivan and the LeadLeo Research Institute in August forecast China's AI agent market will grow from 2.6 trillion yuan in 2025 to 20.1 trillion yuan by 2030, with the B-side contributing about 90% of revenue. In contrast to consumer products, enterprise purchasing involves organization, process and data, making it stickier once agents are deeply embedded. Whether Qianwen Office can convert these enterprises from initial use to sustained payment remains its most immediate test.