Chinese artificial intelligence models had already been outpacing expectations — but they have now pushed the bar even higher. Two local tech groups have recently dropped significant updates. DeepSeek released a dramatically improved V3 model last week, while Alibaba launched a new model in its Qwen series, efficient enough to run on mobile phones.
中國人工智慧(AI)模型的發展已經超出了人們的預期——但它們現在把標準拉得更高了。兩家中國科技集團最近發佈重要更新。深度求索(DeepSeek)上週發佈了經過重大升級的V3；而阿里巴巴(DeepSeek)則發佈了通義千問(Qwen)系列的新模型，其效率足以在手機上運行。
您已閱讀14%（485字），剩餘86%（2987字）包含更多重要資訊，訂閱以繼續探索完整內容，並享受更多專屬服務。
