Google announce a new generative AI audio tool — available now in the Gemini app

Alongside the Pixel 10a announcement, Google has launched a new audio generation tool, Lyria 3, in the Gemini app.

Lyria 3, Google DeepMind’s latest generative music model, is available from today in the Gemini app.  It allows users to create their own music simply by describing an idea, uploading a photo, or even uploading their own inspiration into the Gemini app.

Lyria 3 builds on the previous audio generation by generating lyrics for you based on your prompt, allowing you to control style, vocals, and tempo as you desire and creating more realistic music tracks.  

To use Lyria 3, simply describe the genre, mood, inside joke or memory, and tracks will be created with or without lyrics to fit your description prompt.  You can also upload a photo or video to provide context to Gemini so it can create a track matching the mood in the image or video.

At this stage, the Gemini app will create 30-second tracks with custom cover art, making it simple to share with others by downloading or sharing a link to the song.  Google’s goal with generative-AI-created audio is not a professional track but a “fun, custom soundtrack to your daily life.”

Lyria 3 can also be accessed via YouTube’s Dream Track, which is currently available only in the US but will roll out to other countries in the coming months.  This way, Lyria 3 can be used to create custom soundtracks for your Shorts or uploaded videos.

The audio tracks have SynthID embedded, allowing them to be identified as Google AI-generated content.  Gemini will also receive new tools to help you identify AI-generated content, including audio.

Lyria 3 is available in the Gemini app for all users 18+ in English, German, Spanish, French, Hindi, Japanese, Korean and Portuguese, with plans to expand quality and coverage of more languages.
Google AI Plus, Pro, and Ultra subscribers will enjoy higher limits, although they are not yet listed.  Try it out today at gemini.google.com.