March 26 news.GoogleAnnounce the most advanced under the flagmusic generation model Lyria 3 Pro, a significant improvement over the previous generation, mainly in terms of structure and length of generation。
Google indicates that Lyria 3 Pro is able to more accurately process preludes, lead songs, side songs, bridge segments, etc., and supports the generation of a full track of up to three minutes, which is closer to the end-of-life music that can be used directly。
At the same time, the model does not directly mimic a particular artist's style, and even if the user enters the artist's name in a hint, the model only “as an inspiration”. With regard to training data, the company stated that it used only “the right to use within the terms of the service, the cooperation agreement and applicable law”, but did not disclose more specific data source details. At the same time, Google claims that all audio generated by the model is embedded with an invisible SynthID digital watermark to mark the AI generation。
Currently, the model has landed in several product lines under the Google flag, including Gemini, Google Vids (Workspace Ecology) for business users, the Cloud Platform Vertex AI and the Developer Tool Google AI Studio. In addition, the collaborative music creation tool ProducerAI has been integrated into the model to support the production of songs。