On September 6, 2026, Google officially launched the new Lyria 3.5 music generation model in its Gemini application and API. This model aims to provide more expressive voices and richer arrangement capabilities, and users can access it through channels such as Flow Music, AI Studio, and Google Vids. Google emphasized that Lyria 3.5 used only authorized content during training to ensure compliance with copyright laws.
Google has released the new Lyria 3.5 music generation model in its Gemini application and API. This model promises more expressive voices and richer orchestration, and can be accessed through Flow Music, AI Studio, and Google Vids. Google states that the model is trained only with authorized content.