Google has integrated its advanced Lyria 3 AI music generator into the Gemini app, allowing users to create original 30-second music tracks from simple text prompts. This new feature, currently in beta, expands Gemini's capabilities beyond text, code, and image generation, marking a significant step into the generative audio market.
Key Takeaways
- Google's Gemini app now features Lyria 3, an AI music generation model.
- Users can create 30-second music tracks using text, image, or video prompts.
- The feature is available globally in eight languages to users aged 18 and above.
- AI-generated tracks include custom cover art and embedded SynthID watermarks.
Unleashing Musical Creativity with Gemini
The Lyria 3 model, developed by DeepMind, enables Gemini users to generate music by describing genres, moods, or even specific memories. For instance, a user could request an "Afrobeat track for my mother about the great times we had growing up." The AI can produce instrumental pieces or songs with automatically generated lyrics. Additionally, users can upload images or video clips, and Gemini will compose a track that complements the visual content's vibe.
Google states that the primary goal of these AI-generated tracks is not to produce masterpieces but to offer a fun and unique avenue for self-expression. The generated music will be accompanied by custom cover art created by Google's Nano Banana model, facilitating easier sharing and downloading.
Accessibility and Copyright Considerations
Lyria 3's integration into the Gemini app makes this powerful music generation tool widely accessible, a move that positions Google to compete with emerging AI music platforms. The feature is rolling out globally in English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese, with plans for future language expansion. Access is restricted to Gemini app users aged 18 and older, a measure likely implemented to address copyright and content moderation concerns.
Google has emphasised its commitment to copyright, stating that Lyria 3 is designed for original expression rather than mimicking existing artists. While prompts referencing specific artists will be interpreted as broad creative inspiration, the company has implemented filters to check outputs against existing content and encourages users to report any instances of unintended imitation.
Technical Features and Future Prospects
All music generated by Lyria 3 within Gemini will feature an embedded audio watermark via Google's SynthID technology. This allows users to verify if a piece of audio was created using Google's AI by uploading it back into Gemini. While the current output is limited to 30-second tracks, users with AI Pro and AI Ultra subscriptions will benefit from higher usage limits. Google also plans to bring Lyria 3 to YouTube's Dream Track tool, enabling creators to generate custom AI soundtracks for YouTube Shorts.