Google has added Lyria 3.5, the most advanced music generation model to date, to the Gemini app. The model was previously available through Google’s Flow AI movie-making tool, and now rolled out for all Gemini users, so you can create clean songs, instruments and soundtracks from simple text prompts or even photos.
Lyria 3.5 makes AI-generated music more natural
According to Google, Lyria 3.5 provides richer instrumentation, more expressive vocals and better immediate understanding than previous versions. The model builds entire songs with real verses, choruses, and bridges instead of just looping a short melody clip.
In Gemini, you can now select or describe a genre, choose between vocal or instrumental styles, and decide whether you want a short clip or a longer track. If you’re a beginner, you can use templates to get started and generate background music for a video, custom ringtone, or personalized ringtone.
Note, however, that once a number is created, you cannot edit it in the middle of the prompt. Google is also making Lyria 3.5 available through the Gemini API, allowing developers to integrate AI music generation into their own apps and services.
Lyria 3.5 arrives at a chaotic moment for AI music
AI music has flooded music platforms such as Deezer, where AI-generated songs account for 44% of daily uploads, and many of the streams they download are later found to be fraudulent. That’s why every track created by Lyria includes a SynthID watermark – an invisible marker that marks it as AI-generated content.
Lyria’s safeguards specifically block cloning of an artist’s voice or reproduction of copyrighted lyrics, the latest European court judgment against Sunoa rival AI music company due to copyright issues. However, this judgment is not the final word on the matter. It is still subject to appeal, so the legal picture surrounding music generated by artificial intelligence is still unclear.
\
