The monthly active user count for Tencent WorkBuddy reached 20.97 million visits in June, surpassing the combined portfolio of all Alibaba artificial intelligence products. This situation put pressure on DingTalk's new CEO, Chen Yusen, who decided to launch Qianwen Office within 46 days of consolidating three existing products.
Ma Nan personally praised WorkBuddy during Tencent's quarterly report, calling it the most widely used AI efficiency agent in China. Furthermore, President Martin Lau mentioned WorkBuddy three times in the same call, while Yuanbao was mentioned only once, signaling a strategic priority for Tencent.
WorkBuddy's success is attributed to its integration with WeChat, including QR code login, mini-program integration, and setup without specialized knowledge. Non-technical employees use this tool for creating reports, presentations, coding, and participating in meetings. Tencent captured the ecosystem niche while Alibaba focused on internal competitive races, and ByteDance positioned Kouzi as a developer tool.
However, WorkBuddy faces structural limitations when deeply penetrating the corporate sector. WeCom functions more as a customer communication tool rather than an enterprise workflow operating system. Core business data, approval processes, and organizational rights are handled in DingTalk or Feishu, not WeCom. Qianwen Office, conversely, can directly interact with ERP interfaces, approval mechanisms, and permission systems within the closed DingTalk ecosystem. This allows for tasks such as financial reconciliation requiring ERP inventory data, budget comparison, multi-level approval, and generation of payment orders, all within a single seamless process.
WorkBuddy is forced to extract data from external ERP systems, process it, and then write it back, creating an additional step, delay, and security risk for regulated industries. Meanwhile, DingTalk serves over 26 million enterprise organizations with deep workflow integration, despite the paid conversion rate remaining below 1%. The competitive dynamic is shifting from consumer feature competition to the depth of integration into corporate workflows. WorkBuddy's strategy, based on expanding the skills ecosystem and gradually introducing enterprise functions through partnerships with ZTE, Meituan, and Ping An Bank, represents a long-term peripheral approach. In contrast, Qianwen Office's strategy, focused on deep integration with DingTalk, aims for the position of an enterprise operating system, where AI becomes an integral part of every workflow—from expense reimbursement to hiring and sales analysis.
The battle line runs between the advantage of WorkBuddy's consumer ecosystem and the depth of DingTalk's corporate integration. The ultimate winner will depend on which approach converts more effectively into paid corporate subscriptions. Tencent faces the challenge of gaining enterprise trust without an enterprise operating system infrastructure, whereas DingTalk struggles with the task of converting its massive base of free users into paying customers. Both approaches have yet to prove their effectiveness at scale for office AI agents, making the desktop office AI market the most significant strategic battleground in Chinese consumer-enterprise technology.
The city of Dongguan, previously known as a major center for low-cost assembly manufacturing in China, is transforming into a global hub for intelligent production thanks to the boom in artificial intelligence hardware. In 2025, Dongguan produced half of the world's smart glasses volume, supplying 870 million units globally, with over 400 million manufactured in Dongguan factories, representing a 322% increase compared to the previous year.
Meta holds a dominant position in the global market, controlling 85.2% of the market, with its main OEM factories located precisely in Dongguan. This large-scale transformation is enshrined in the Global Intelligent Manufacturing Center plan, approved in May 2026, where AI-based production is defined as a priority among six key initiatives.
Currently, the city operates 24,000 industrial enterprises, over 10,000 high-tech firms, and 195 million market entities. Industrial output reached 2.6 trillion yuan in 2025, ranking second in Guangdong province.
The AI hardware ecosystem benefits from an exceptional supply chain density. Within the city area of 2,460 square kilometers, 90% of key components for smart glasses can be found within a one-hour radius. A single supply chain concentrates optics from Depute and Gaowei, acoustics from Goertek and Jiahe, structural elements from Changying and Lens, storage from Biwin, and batteries from New Energy Technology.
The OPPO factory in Dongguan Bei achieves 98% automation on assembly lines, which is one of the highest rates in Chinese smartphone manufacturing, where material transportation and assembly are performed using robotic manipulators and AGVs.
The National AI Pilot Base, the only national base for consumer AI mobile terminals, provides chip adaptation, HarmonyOS integration, and AI model testing services for small and medium-sized enterprises that cannot afford their own verification production lines, thereby eliminating a critical bottleneck for over 30,000 firms involved in intelligent equipment in the city.
Companies such as Slike have undergone significant changes: they transformed from a USB flash drive workshop established in 2005 into an AI smart glasses manufacturer, supplying over 100 million units annually as an exclusive supplier for MINISO AI translation glasses and an ODM provider for Lenovo and Coolpad.
Huayan New Materials, the world's largest producer of AI smart glasses hinges, supplies 60% of Meta's hinge orders, and its production capacity for 2025 corresponds to 7 million pairs. Siqian New Materials produces ultra-thin synthetic graphite film for heat dissipation with a thickness of 0.05 mm.
In 2025, the city launched China's first urban center for large-scale production models, adding over 10,000 petabytes of available intelligent computing power. Furthermore, a computing industrial island is under construction on Shazhou Nichou Island with investments of 100 billion yuan over five years, aimed at achieving a server industry volume of 500 billion yuan by 2030. Dongguan already produces nearly 60 million servers annually, accounting for one-tenth of China's servers, with 95% of components sourced locally. The city has also implemented the first XR policy specific to prefectural cities in China, offering one-time grants of up to 60 million yuan for relevant OEM enterprises.
Tencent WorkBuddy reached 20.97 million monthly visits in June 2026, surpassing the combined figures of ByteDance Trae (12.79 million) and Alibaba QoderWork (7.88 million). This success was achieved just four months after its launch in March 2026, driven by the use of the OpenClaw ecosystem and integration with corporate WeChat.
In June 2026, according to an Analysys ranking, Tencent WorkBuddy demonstrated 20.97 million monthly visits, ahead of ByteDance Trae at 12.79 million and Alibaba QoderWork at 7.88 million combined. The product launched in March 2026, showing 8.85 million visits in its first month and an impressive 831% growth compared to the previous month, which triggered a tenfold capacity expansion in the first week alone. This rapid ascent occurred while industry attention was focused on model competitions such as Kimi K3, GLM-5.2, and DeepSeek V4, allowing Tencent to capture the desktop AI office agent category without a direct competitive response.
The foundation of WorkBuddy's success is the OpenClaw ecosystem—an open-source framework that had 280,000 stars on GitHub and over 5,700 plugins before Tencent's integration. On March 9th, Tencent simultaneously released three products: QClaw for private users, Lighthouse for developers, and WorkBuddy for enterprises. WorkBuddy is fully compatible with the OpenClaw skill set, includes over 20 pre-installed skill packages, supports parallel task decomposition using multiple agents, and provides model flexibility, allowing interchangeable use of Hunyuan, DeepSeek, GLM, Kimi, and MiniMax. The corporate integration chain covers input via WeChat and WeCom, computation on Tencent Cloud, and local device permissions through PC Manager, forming a closed loop that competitors find difficult to replicate due to the breadth of the ecosystem.
The core infrastructure of the Hunyuan model was rebuilt in early 2026. The HY3 preview held the top spot in the weekly OpenRouter ranking for three consecutive weeks in April before being overtaken by DeepSeek V4 Flash. The official release on July 6 reduced hallucination rates by 44% and increased WorkBuddy's task resolution rate from 72% to 90%. With pricing at 1 yuan for input and 4 yuan for output per million tokens, the product remains competitive. Tencent reported active user retention above 60% and paid user retention above 80%, although these figures combine metrics from CodeBuddy and WorkBuddy. CFO James Mitchell declined to set an ARR target, noting that demand changes are happening too quickly for meaningful 12-month forecasting.
The competitive battle is transforming the AI office agent market from a competition of model performance to a competition of ecosystem and distribution. WorkBuddy's advantage stems not from superior model quality, but from existing enterprise relationships within WeCom, WeChat consumer reach, and the OpenClaw plugin ecosystem, which reduces user adoption time from days to minutes. In response, Alibaba consolidated its products into Qianwen Office under new CEO DingTalk Chen Yusen, while ByteDance restructured Doubao as its office agent brand. Nevertheless, both companies face a structural challenge: WorkBuddy is integrated into the WeChat corporate ecosystem, which is already used for business communication in China. The question remains whether network effects for group collaboration can form around WorkBuddy, comparable to the network effects of consumer WeChat, potentially securing it a sustainable position that competitors will find difficult to disrupt.
According to a report from ShanghaiTech University CEISD, based on the analysis of 65,200 individual tasks across 1,267 different professions, artificial intelligence in China is not merely eliminating jobs but fundamentally transforming them. The analysis covered over 700 million job postings on five major platforms between 2022 and 2026.
The central finding of the study is that approximately 21 million positions have been reshaped through a combination of tasks, accounting for 18.6% of the total number of open vacancies. The methodology shifted focus from job titles to the composition of tasks performed, revealing that high susceptibility to AI does not always equate to unemployment risk.
The study compared Java developers and customer support representatives. While script processing and ticket entry in customer support can be handled by large language models (LLMs) with minimal human involvement, Java developers retain high-level tasks such as architectural design, fault diagnosis, and on-site decision-making, which drives salary growth for senior engineers despite the automation of junior coding tasks.
Reductions most severely affect entry-level and mid-level positions. Among the 33.2% of roles classified as task combinations with high AI exposure, 26.6% fall into the zone of replacement pressure with high exposure and low complementarity, meaning one of four online vacancies is subject to structural adjustment caused by AI. The impact is unevenly distributed: entry-level workers earning between 5,000–8,000 yuan with 1 to 3 years of experience face the greatest reduction, while positions above 12,000 yuan show full recovery, and roles exceeding 50,000 yuan demonstrate positive momentum.
The traditional training process for junior specialists, which includes document processing, data entry, and standardized communication, is disappearing as these functions are absorbed by LLMs. Contrary to expectations of a mass emergence of new professions, the report notes only a minimal increase in job titles since 2023. A slight increase is observed only in the titles 'LLM algorithm engineer' and 'robotics engineer,' while traditional roles in recommendation, risk control, and SLAM continue to decline. The fastest-growing are not new professions, but existing roles transformed through task recombination. The defining new employment type is the composite closed loop, where employers increasingly prefer candidates capable of managing the entire cycle—from requirement definition to execution, verification, and reporting—instead of narrowly specialized employees.
The market correction has gone through three stages: a 2.6% reduction in 2024 during tool experimentation, a panicked surge of 30.1% in 2025, and a rational stabilization at 16.9% in 2026. Those who combine AI complementarity with opportunities for on-site work and risk responsibility gain an advantage, challenging the traditional entry path into the labor market for recent graduates.
