Google is expanding the creative tools within its Gemini app with the rollout of Lyria 3 Ai, a new generative music model developed by Google DeepMind. Currently available in beta, the feature allows users to generate short music tracks using simple text prompts or by uploading photos and videos.
With Lyria 3 Ai, users can describe an idea, mood, memory, or inside joke, and Gemini will turn it into a 30-second song, complete with AI-generated lyrics and cover art. The tool is designed for quick, informal creativity rather than professional music production, positioning it as a lightweight way to add a personal soundtrack to everyday moments.
More Control and Improved Output
Compared to earlier versions, Lyria 3 Ai introduces several upgrades. Users no longer need to write their own lyrics, as these are generated automatically. The model also offers more control over musical elements such as tempo, vocal style, and genre. According to Google, the system now produces more detailed and musically complex tracks, with improved realism.
The feature supports two main modes:
- Text to Track: Users can describe a genre, emotion, or memory to generate a song. For example, a nostalgic prompt about family memories can be turned into an Afrobeat-inspired track.
- Media to Track: Photos or videos can be uploaded, which Gemini then uses as creative input for composing lyrics and music that reflect the mood of the visuals.
Each track is generated alongside custom artwork, making it easy to download or share directly from the app.
Integration With YouTube Shorts
Lyria 3 Ai is also being made available through YouTube’s Dream Track feature, initially in the US and now expanding to other regions. For Shorts creators, the model can generate custom background music or lyrical snippets, offering more flexibility in soundtrack creation.
Verification and Responsible Use
All music generated through Gemini is embedded with SynthID, Google’s digital watermarking system for AI content. This allows users to verify whether audio, images, or videos were created using Google’s tools. The company has also added audio detection capabilities to Gemini’s existing verification features.
Google says Lyria 3 Ai has been developed in collaboration with the music community, with attention to copyright and licensing concerns. The system is designed to create original material rather than imitate specific artists. If a prompt references a particular musician, Gemini interprets this as general stylistic inspiration rather than direct emulation. Outputs are also checked against existing content, and users can report potential rights violations.
Availability
Lyria 3 Ai is now available in the Gemini app for users aged 18 and above in multiple languages, including English, Hindi, Japanese, Korean, French, German, Spanish, and Portuguese. Desktop access is live, with mobile availability rolling out gradually. Subscribers to Google’s AI Plus, Pro, and Ultra plans receive higher usage limits.
With this update, Google is positioning Gemini’s music tools as a casual, creative layer within its broader AI ecosystem, aimed less at replacing musicians and more at enabling everyday users to experiment with personalised soundtracks.
Try it out today at gemini.google.com.
