Google has officially added Lyria 3.5, its most advanced music generation model yet, to the Gemini app. The technology was originally introduced a little over a month ago. At the time of its initial announcement, the model was only available in Google Flow Music, which serves as the company's AI filmmaking tool.
Now, the AI model is accessible in a couple more places. Starting today, Lyria 3.5 is available in both the Gemini app and the Gemini API. This wider rollout brings the model to web platforms, mobile devices, and Flow Music simultaneously.
Richer Tracks and API Access
With the launch of Lyria 3.5 on the Gemini app, users can easily generate polished songs, instrumentals, and custom soundtracks. The model allows people to create these tracks from simple text prompts or even by uploading photos. Additionally, Google says you will now be able to choose between short and long tracks depending on your needs.
Users can pick or describe their desired genre while choosing between vocal or instrumental styles. The company states that Lyria 3.5 makes AI-generated music sound more natural by delivering richer arrangements, more expressive vocals, and better prompt comprehension than previous versions.
As a brief overview of what the upgrade brings to the table, the model was designed to produce richer and more complex melodic structures. It better adheres to prompts, is highly structurally aware, and noticeably improves upon overall pronunciation.
According to Google, the updated vocals are supposed to be more realistic and emotionally nuanced. To achieve this, the model now builds out full songs complete with actual verses, choruses, and bridges. This marks a major shift from earlier tools that simply looped a short melodic clip.
If you are a beginner, you can use newly added templates to get inspiration and start generating audio quickly. These templates can help you generate background music for a video, a custom jingle, or a personalized ringtone. However, users must keep in mind that once a track is generated, you cannot edit it mid-prompt.
Google is also making Lyria 3.5 available through the Gemini API, allowing developers to integrate AI music generation directly into their own apps and services. As part of this rollout, developers now have access to the model in Google AI Studio and Google Vids.
The Rising Tide of AI Music
The widespread rollout of Lyria 3.5 arrives at a chaotic moment for the music industry and AI content. AI music has aggressively flooded music platforms like Deezer. On that specific platform, AI-generated songs now make up a staggering 44% of daily uploads.
Furthermore, a large share of the streams these generated tracks have pulled in were later found to be fraudulent. To help mitigate this issue, every track Lyria makes carries a SynthID watermark. This acts as an invisible marker meant to flag the audio as AI-generated content to platforms and listeners.