Tencent's WeChat develops proprietary AI model amid heavy AI spending
Tencent spent $3.3B on R&D in Q1, prioritizing internal AI over cloud GPU leasing.
Deep Dive
Tencent is integrating AI into its businesses including WeChat, Yuanbao, advertising, games, and more. In Q1 2026, revenue hit $28.8B (up 9% YoY), but heavy AI investment slowed profit growth to single digits. R&D spending rose 19% to $3.3B, and capex reached $4.7B (up 16%). Tencent launched agent products (WorkBuddy, QClaw) and open-sourced Hy3 Preview. The company prioritizes internal AI services—like those for WeChat and advertising—over leasing GPU capacity, which has limited cloud business growth.
Key Points
- Tencent spent $3.3B on R&D and $4.7B on capex in Q1 2026, driving profit growth down from potential 17% to 9%.
- Hy3 Preview open-sourced in April; token usage grew over 10x, with coding and agent scenarios up 16.5x.
- Tencent prioritizes internal AI use (including WeChat, Yuanbao, advertising, games) over cloud GPU leasing.
Why It Matters
WeChat's proprietary AI model could reshape China's messaging ecosystem, but heavy spending pressures near-term profits.