Google 发布 Lyria 3.5 音乐模型，直接接入 Gemini 应用与生产力生态

My Analysis of Google's Lyria 3.5 Launch

Okay, so here's what I've got. Google has rolled out its Lyria 3.5 music model. It's available in the Gemini app and also via API, which means developers can access it directly. The model's supposed to offer improved vocal expressiveness and richer arrangements – that's a key selling point. It's also integrated into Flow Music, AI Studio, and Google Vids, broadening its accessibility. The thing that catches my attention from a regulatory/ethical perspective is that Google is emphasizing that this was trained only on licensed content. That's a strong statement, and likely a response to the ongoing discussions around AI copyright and content usage. And finally, the article reporting on this launch first appeared on The Decoder.

谷歌已在 Gemini 应用中并通过 API 推出了其 Lyria 3.5 音乐模型。该模型带来了表现力更强的人声和更丰富的编曲，同时也可通过 Flow Music、AI Studio 和 Google Vids 使用。谷歌表示，该模型仅基于获得授权的内容进行训练。文章《谷歌通过全新 Lyria 3.5 模型将 AI 音乐生成直接引入 Gemini 应用》首发于 The Decoder。