Why WeChat's AI entry doesn't use Tencent's Hunyuan: data security keeps it on a small in-house model
dotey · x · 2026-10-20
An analysis of WeChat's AI stack: its AI entry point doesn't use Tencent's Hunyuan LLM but a small in-house model called Xiaowei, chosen purely for data security — a national-scale app won't hand user data to any internal team. WeChat thus runs three model systems (Xiaowei, Yuanbao, Hunyuan), with most Tencent teams expected to adopt Hunyuan while WeChat keeps building Xiaowei in-house.
More from Companies & People
- Amazon blocks AI shopping agents while Walmart opens its site to them with spend controls — venturetwins · 2026-10-23
- Commenters speculate Skydance could wire Grok into its entire Hollywood production pipeline — MickeySteamboat · 2026-10-23
- Anthropic, OpenAI and Musk pledge over 3x the Gates Foundation, each with a different model — rohanpaul_ai · 2026-10-23
- CU Boulder's ATLAS Institute Opens Tenure-Track Robotics Professorship — ryosuzk · 2026-10-23
- Andreessen: AI-Native Startups vs Incumbents Show Widest Perception Gap He's Ever Seen — juanbenet · 2026-10-23
- Stanford opens up its Principles of Robot Autonomy teaching materials for free — GioeleZardini · 2026-10-23