腾讯混元的Hy实时语音模型能够从半句语音中重建意图,决定何时搜索网络,并可应要求唱歌。它已在元宝语音通话中上线。Tencent Hunyuan's Hy Realtime voice model rebuilds intent from half-finished speech, decides when to search the web and sings on request. It is live in Yuanbao voice calls.