[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tag-多模態-ai":3},{"tag":4,"articles":11,"peer_article_count":12},{"id":5,"name":6,"slug":7,"article_count":8,"description_zh":9,"description_en":10},"2e479ff2-c252-4b0a-8a0c-627056219240","多模態 AI","多模態-ai",3,"多模態 AI 結合文字、語音、影像與工具呼叫，讓模型不只會回答，也能看圖、聽聲、即時互動並執行任務。從即時語音代理到視覺寫碼與多代理工作流，這類模型正重塑應用介面與開發方式。","Multimodal AI combines text, speech, images, and tool use so models can interpret inputs and act in real time. It matters for live voice agents, vision-enabled coding, and agent workflows that link perception with execution.",[],5]