The rise of office‑focused AI Agents has reshaped enterprise productivity software. As AI evolves from simple chat‑answer assistants toward actionable agents that can operate software, generate spreadsheets and complete multi‑step business workflows, major tech companies are racing to roll out their representative products. Tencent’s WorkBuddy has quickly become one of the most watched entries in this segment. Internally nicknamed the “new crown prince”, this office‑agent product has achieved explosive user growth in a short period. This article traces its origin, product iteration, operational milestones, competitive landscape and upcoming engineering and commercial challenges, while analyzing where the office‑agent track is heading.
Inception and Early Iteration of WorkBuddy
Back in January 2026, Wang Chengjie, a Tencent Cloud product manager, saw potential in CodeBuddy, an internal AI coding assistant. CodeBuddy was heavily adopted internally, with more than 90 percent of Tencent engineers relying on it for AI‑assisted code generation. The product team conceived an idea: extend its AI capability beyond coding to general office scenarios, including report drafting, spreadsheet processing and routine office tasks. Three core team members built the initial prototype within a single weekend, giving birth to WorkBuddy.
At that point in January 2026, WorkBuddy existed only as an experimental internal prototype supported by a small team of around a dozen staff. On January 19, Wang Chengjie delivered a demonstration report to senior leadership. The demo showed WorkBuddy finishing complex multi‑step office tasks in roughly ten minutes. Senior management quickly green‑lit full‑company internal testing. Non‑technical employees started using it for data analysis and document‑related work. Within two months, more than 2000 non‑technical staff inside Tencent participated in the internal trial.
The market environment outside Tencent accelerated the product’s public launch. OpenClaw went viral across the developer community, popularizing the concept of AI agents that can simulate human desktop operations. The industry reached a consensus: AI is shifting from conversational Q&A toward actionable execution. Under this competitive pressure, Tencent could no longer afford long‑drawn‑out polishing cycles. On March 9, less than two months after internal testing kicked off, WorkBuddy was officially released to the public. Its release day request volume hit dozens of times of CodeBuddy’s baseline traffic, and engineering teams faced urgent pressure to scale up infrastructure capacity. On March 12, ahead of schedule, it opened direct integration access with WeChat.
During the same window, competing products flooded the market. ByteDance launched its Volcano Ark Agent offering. Alibaba rolled out Qianwen‑based agent capabilities. DingTalk also released comparable office‑agent solutions. The whole industry entered a “dragon‑tiger‑showdown” phase, with every major internet player pushing similar agent‑style office products.
Rapid Growth and Internal Strategic Upgrade
WorkBuddy posted impressive growth metrics after launch. In March, its PC‑side monthly active users reached 8.85 million. By June, monthly PC‑side active users surged to 20.97 million, securing first place among domestic comparable office‑agent products. Its user volume exceeded the sum of the second‑ and third‑ranked competitors combined.
Parallel to user‑number expansion, WorkBuddy’s internal strategic weight rose sharply. Ma Huateng gave repeated public affirmation. The core team expanded from dozens of members to more than one hundred. In May, the team relocated to Tencent Cloud headquarters. Liu Jing, the department lead, received a promotion. Inside Tencent, employees began referring to WorkBuddy as the company’s “new crown prince”, reflecting high‑level expectations for this product.
In July, parts of QClaw, Tencent’s earlier agent‑oriented product, were merged into the WorkBuddy system. WorkBuddy was formally positioned as Tencent’s flagship office‑agent product suite. It established connections with Hunyuan large‑model series and other internal Tencent business modules. Twenty‑plus years of Tencent’s accumulated enterprise user resources became a powerful moat for WorkBuddy. In Q1 earnings call materials, Ma Huateng described WorkBuddy as “China’s most widely‑adopted efficiency‑oriented AI agent service”.
Tencent invested heavily in user acquisition. Offline‑and‑online advertising spending for WorkBuddy exceeded 100 million RMB in the first half of the year. To lower access barriers, it supported multiple third‑party large‑model back‑ends and rolled out registration bonus incentive campaigns. The core operational target was set toward daily active‑user growth.
When integrating heterogeneous large‑model back‑ends, enterprises often face authentication management, traffic throttling and model fallback complexity. An API gateway can simplify multi‑model access architecture. 4sapi provides unified routing capability for mixed‑model workloads, reducing repetitive integration work for engineering teams building agent applications.
Intense Competition Within the Office‑Agent Track
The whole office‑agent market is expanding at high speed. According to iMedia Research statistics, China’s AI intelligent‑agent market size reached 80.4 billion RMB in 2025, marking a year‑over‑year growth rate of 123.2 percent. Projections show the market will hit 696.8 billion RMB by 2030. This huge growth potential attracts heavyweight participants from every major internet firm.
ByteDance expands its developer‑oriented tools toward workplace scenarios. It reorganizes internal teams to deliver MaaS and SaaS sales workflows. Alibaba integrates Qianwen Agent capabilities inside DingTalk. New DingTalk CEO Chen Yuxuan takes charge of pushing agent‑feature rollouts. Baidu launches dedicated office‑agent modules on its Qianfan platform. Besides internet giants, software and hardware vendors also join the competition. They upgrade strategic business architectures and push agent‑native products to grab enterprise‑service market share.
Third‑party research data collected as of mid‑2026 records 17 mainstream desktop‑side native AI office‑agent products on the Chinese market. Competition is extremely fierce. No single player has locked in absolute dominance. User experience, ecosystem compatibility, cost of invocation and enterprise‑level security governance become key dimensions that separate competing products.
Agent‑native office applications impose strict requirements on backend API stability. Every multi‑step agent task may trigger dozens of consecutive model calls. Unstable interfaces or rate‑limit interruptions can break entire workflow executions. Therefore, traffic scheduling, error retry and multi‑model fallback mechanisms are indispensable for production‑grade agent services.
Challenges Lying Ahead for WorkBuddy
By July 2026, WorkBuddy accumulated more than 340 built‑in expert modules. Coupled with the SkillHub marketplace hosting over ten thousand skills, it formed a sizable capability pool. End‑users can input natural‑language instructions to trigger complete end‑to‑end office workflows. It supports three‑terminal synchronization covering desktop, web and mobile clients, and delivers remote desktop control capability. In the domestic desktop‑native AI office‑agent segment, WorkBuddy occupies the top market position.
Nevertheless, multiple tough challenges remain. Tencent keeps pouring capital into this product. Q2 financial filings show increased capital expenditure, with substantial bond‑financing funds allocated to AI‑related projects, and WorkBuddy ranks high on the investment priority list. Mitchell, Tencent’s chief strategy officer, pointed out that compute‑resource allocation will tilt heavily toward agent‑related business. Large existing user base and ecological resources are WorkBuddy’s core strengths, yet real‑world commercial conversion still needs further validation.
The Chinese enterprise‑software market follows a clear logic: enterprises will only pay for tangible business returns. Scale‑driven free‑user growth cannot sustain long‑term operation without viable monetization paths. Industry statistics show that the total monthly active‑user base for domestic AI office desktop‑agent products stands at roughly 30 million, with estimated 20‑fold room for future user‑base expansion. But converting free users into paying enterprise clients remains the biggest hurdle for WorkBuddy and its competitors.
Technical risks also persist. Desktop‑agent products rely heavily on LLM reasoning accuracy. Hallucinations, logical deviation and mis‑operation of local files may cause business losses for corporate users. Enterprises raise strict requirements for permission granularity, operation auditing and data privacy. Even if model capability keeps improving, reliability and safety constraints will slow down large‑scale enterprise adoption. WorkBuddy needs to balance rapid feature iteration with enterprise‑grade stability and compliance standards.
Looking at the whole industry, the office‑agent track is still in the growth phase. No player has delivered fully‑mature, universally‑accepted commercial solutions. Product form, pricing mode and enterprise service standards are still being explored. WorkBuddy holds first‑mover advantages in user volume and ecosystem resources, but it still needs to prove its commercial value.
Conclusion
WorkBuddy’s rapid rise mirrors the explosive development of China’s office‑agent industry. Starting from a small internal prototype, it grew into a top‑tier public product with tens of millions of monthly‑active users within half a year. Supported by Tencent’s computing resources, user base and channel resources, it occupies a favorable position in the fierce agent competition.
Still, high user traffic does not equal commercial success. The core test ahead lies in transforming active‑user scale into measurable business value for enterprise customers. Technical stability, permission security, skill‑ecosystem construction and monetization mechanism design will decide which participant can finally win out in the office‑agent arena. As agent technology continues to evolve, the whole industry will keep observing whether WorkBuddy can live up to its high internal expectations.
Learn more：https://4sapi.com
