Tencent's Midday Surge of Over 5% Marks Narrative Reversal: WorkBuddy User Base Grows by Two Orders of Magnitude, WeChat AI Assistant "Xiao Wei" Expands Boundaries
BlockBeats News, July 6th, according to Bitget market data, Tencent Holdings' stock price in Hong Kong rose more than 5% during intraday trading today, reaching 453 Hong Kong dollars. Tencent, once criticized for being "slow in AI," is now experiencing a narrative reversal. Pony Ma admitted at the shareholders' meeting in May that "a year ago, I thought I had boarded the ship, but later found that the ship was leaking. Now that I've stepped on board, I still can't sit down." However, in just a few months, WorkBuddy's monthly active users exceeded 20 million, daily active users exceeded 13 million, DAU/MAU ratio reached 65%-75%, with stickiness comparable to Slack-level office tools. In three months, 43 versions were released, and updates continued as usual during the May 1st holiday. Tencent internally positions it as the third product with a phenomenal impact after QQ and WeChat.
The key to WorkBuddy's success lies in two barriers: first, extremely simple installation—desktop client plus WeChat mini-program. After WeChat authorization, commands can be sent from the phone and automatically executed on the PC, completely bypassing the high threshold of command-line deployment for similar foreign products. Second, Tencent's full-stack ecosystem was activated in the Agent era—having integrated Tencent Docs, Tencent Meeting, Enterprise WeChat, and more than 30 external tools. The SkillHub skill library has expanded from 79,000 to 790,000. Simply sending last week's sales data in a PPT to attending colleagues can automatically fetch data, generate files, and distribute them. The leap from a chatbot to a tool that helps you work has transformed a fragmented toolchain into a moat. The fact that Tencent, for the first time in its history, has promoted a product at a grassroots level also reflects its true determination to target the B-end market.
Further imagination comes from WeChat AI. WeChat AI assistant "Xiao Wei" started a small-scale grayscale test on June 20th, supporting sending messages via voice or text, posting on Moments, making appointments through mini-programs for registration and ordering takeout, and even generating mini-program prototypes through natural language. This is the first time the AI application, with over 1.43 billion monthly active users, has entered the intelligent body race.
JPMorgan Chase estimates, based on probability weighting, that WeChat AI Agent could bring incremental revenue of about 126 billion yuan and incremental operating profit of about 88 billion yuan by 2030; however, Goldman Sachs is cautious about three major concerns: the self-developed WeLM model operating in parallel with MixNet may lead to resource duplication, the cost of inference after full promotion may erode 5%-17% of adjusted operating profit in Q4, and the short-term monetization path is not yet clear. The core of the disagreement between the two institutions lies in the pace: JPMorgan Chase believes that the public beta has transformed WeChat AI from an "abstract option" to a "traceable milestone," while Goldman Sachs believes that the short-term cost and benefit asymmetry requires more quarters of validation.
Regarding the underlying model, MixNet 3 Preview has been integrated into 131 Tencent products. The Token call volume has increased by 10 times compared to the second generation, ranking first in usage on the OpenRouter platform. However, Pony Ma's judgment of not being able to sit still even after stepping on the board still holds. The ecosystem is the natural soil for Tencent to act as an Agent, but WeChat AI needs to deal with a complex proposition of transactions, advertisements, payments, merchants, fairness, and regulation. Its ceiling is higher, but the friction is also greater.
