Google brings AI music generation directly into the Gemini app with its new Lyria 3.5 model

What it does

Google launched the Lyria 3.5 music model inside its Gemini app and opened access through its API as well. This model supports AI-generated music with more expressive vocals and richer arrangements. Google also integrated the model into Flow Music, AI Studio, and Google Vids, widening its availability across creative tools. The training used only fully licensed content, aiming to avoid legal or ethical issues around copyright.

Why it matters

Having AI music generation built directly into the Gemini app means content creators, marketers, and hobbyists can produce higher-quality, customized soundtracks without needing extensive musical skill or separate software. The focus on expressive vocals and better arrangements sets this apart from many prior AI music models that often felt robotic or repetitive. By restricting training data to licensed music, Google addresses one of the key pain points in AI music rights, which can deter businesses from adopting AI-generated audio.

Who it is for

This capability targets digital content creators who need quick, on-demand music generation for videos, apps, podcasts, and games. Small businesses and startups can leverage AI music to reduce production costs usually tied to licensing or hiring musicians. Developers and agencies using the Gemini API can build custom audio experiences or embed dynamic soundtracks in their products. While pros might still prefer traditional production, Lyria 3.5 narrows that gap for many use cases.

The catch

The quality and expressiveness improvements sound promising, but hearing AI vocals that truly compete with human singers will be the real test. Relying on licensed content limits training data variety, which may restrict style diversity. Also, it remains to be seen how expensive accessing the API will be at scale and if Google will enforce usage limits or licensing fees. Users need to weigh the costs against the convenience and speed benefits.

What to watch next

Watch for early user feedback on the Lyria 3.5 model’s vocal naturalness and arrangement complexity in real-world creative projects. Pricing details and API usage policies will be key for businesses considering integration. Google’s broader AI strategy in music will be clearer as it rolls this tech into more products or partners it with music services. Competitor moves from rivals like OpenAI or Meta will test whether Google’s focus on licensed data and integrated apps gives it an edge or slows innovation.

AI Quick Briefs Editorial Desk