Tencent Urgently Needs to Launch an AI “Plan B” — But Does Plan B Exist?

Market wake-up call

Tencent (腾讯) has just received a stark market warning. It has been reported that in the ten trading days before the Spring Festival the company’s shares fell about 14%, versus declines of roughly 4% and 8% for the Hang Seng and Hang Seng Tech indexes respectively. Investors appear to be punishing one thing above all: weak progress on consumer-facing AI applications. According to QuestMobile figures reported by Huxiu, daily active users (DAU) on competing apps were far ahead as of Feb. 7 — 千问 about 73.5 million, 豆包 about 78.7 million, and Tencent’s 元宝 only about 18.3 million.

Why Tencent’s shortfall matters

Tencent is not just another tech firm for Western readers: it owns WeChat, the dominant social platform for over a billion users, plus a huge gaming and ad business that funds new bets. So if Tencent can’t translate its social moat into AI-driven social features, the implications are structural — not just a product miss. It has been reported that Tencent limited the viral spread of 元宝 invitation links and红包 (red-packet) mechanics inside WeChat; such curbs reflect the fine line the company must walk between internal product promotion, regulatory scrutiny and fair competition across its ecosystem. Even with preferential treatment inside WeChat, analysts say catching up to rivals would be an uphill battle.

Competition and investment gaps

Meanwhile competitors are moving fast. ByteDance (字节跳动) has pushed video-model upgrades such as Seedancer 2.0, and Alibaba (阿里巴巴) has open‑sourced Qwen 3.5 — moves that have strategic and global resonance. Tencent’s base-model efforts, notably 混元 (its foundational model project), have lagged in resources and public progress; it has been reported that the recent hire of Yao Shunyu sparked renewed confidence, but big-model R&D is a multi‑year undertaking. Add geopolitical constraints — U.S. export controls and restrictions on advanced AI chips — and the race for model scale and infrastructure becomes even more fraught for all Chinese players.

The choice ahead: overhaul or fall behind

History offers a template: Google reportedly declared a company-wide “code red” after ChatGPT’s emergence and then rapidly reorganized and invested to rebuild its lead. Tencent still has the cash — gaming and video-ad revenues provide ample war chest — but time is shorter. To mount a credible Plan B, it must simultaneously accelerate base‑model investments, product development and hands‑on user operations; doing one without the others will likely fail. The question is blunt: can Tencent move quickly and deeply enough, or will its social advantage erode as AI rewrites how people interact online?