NextFin News — Tencent is not trying to out-build ByteDance app for app. It is trying to make the next wave of attention pay rent inside software people already open every day. A short-drama service running as a WeChat mini program—known in the market as Green Bubble—reached about 89.7 million weekly active users in the first week of October, up roughly 44% from early September. Tencent did not launch a new consumer App to get there. It used WeChat’s distribution, social sharing, and payments.

That path looks slow next to ByteDance’s scoreboard and fast next to the cost of buying installs. Hongguo, ByteDance’s dedicated short-drama App, still operates on another scale—on the order of 368 million monthly users and 168 million daily users in recent third-party reads. It also took years of product work and heavy user acquisition to train the market. Green Bubble’s weekly base is smaller. The efficiency of adding those users inside WeChat is what competitors are watching.

ByteDance’s record is a string of independent products—news feeds, short video, short drama, consumer AI—each fighting for home-screen space and each fed by recommendation and paid growth. Short drama’s breakout followed that template: large content supply, aggressive marketing, a brand users seek by name.

Tencent’s answer in drama has been to avoid a second download. Open WeChat, tap a mini program, watch, share into a group or Moments, pay with WeChat Pay if a title locks. Acquisition is not free—tens of thousands of ad creatives still ran in a recent week—but it is a different order of magnitude from the multi-billion-yuan install campaigns associated with standalone drama apps. The bet is that social forwarding and zero new-install friction can harvest demand ByteDance already proved exists.

The same pattern shows up in work software. Tencent’s WorkBuddy, an AI helper aimed at office tasks, posted about 32.4 million monthly active users in the third quarter, leading a narrow category of office agents after extremely steep sequential growth. It sits against enterprise WeChat, Tencent Docs, and related workflows rather than only against general chatbots that win consumer mindshare through advertising. ByteDance’s Doubao remains the more visible mass-market AI App. Tencent is optimizing for being present when documents and work chats are already open.

Attention metrics favor ByteDance in many public comparisons. Profitability frames the counterpoint. Tencent reported that after stripping the revenue, costs, and expenses of newer AI products—including its model stack and assistants such as Yuanbao, CodeBuddy, WorkBuddy, and Xiaowei—its second-quarter non-IFRS operating margin was 42%, up three points from a year earlier. That does not mean AI is cheap. It means the core franchise, measured that way, still thickened while experiments run.

ByteDance’s first-half 2026 revenue has been discussed in the market around $120 billion, with net margin near 17% under heavy spending on data centers and chips, and a private valuation often cited near $600 billion. Different companies, different disclosure, different stages of build-out. The strategic contrast still holds in outline: one model spends heavily to own the open screen; the other spends to deepen an ecosystem where sharing, payment, and work already clear through its pipes. Tencent’s own capital expenditure remains large—on the order of 53 billion yuan in the second quarter, with full-year guidance discussed above 200 billion yuan—but more of that spend is described as platform infrastructure than as perpetual install subsidies for a single new App.

On time spent, daily actives, and consumer AI awareness, ByteDance has shaped the last several years. On the cost of monetizing new content formats and on operating profit in the core business, Tencent is making a quieter case. Green Bubble does not overtake Hongguo by existing. WorkBuddy does not displace Doubao by leading a narrower office ranking. Together they show a consistent choice: put short drama and AI where WeChat and work accounts already are, rather than demand another icon.

Screen time still matters. It is no longer the only number that decides who extracts value from the next format. Infrastructure that already sits under sharing and payment can turn a proven category into incremental yield—without implying that the contest for pure attention is over.