Gemini App Introduces AI Music Generation with Lyria 3
Key Points
- AI music generation from text or images
- 30-second tracks with auto-generated lyrics
- SynthID watermarking for content verification
Summary
The Gemini app now features Lyria 3, Google DeepMind's latest generative music model, enabling users to create 30-second custom music tracks from text prompts or uploaded images. This beta feature includes automatic lyric generation, custom cover art, and SynthID watermarking for AI content identification.
Key Points
- Text-to-Music Generation: Create tracks by describing genre, mood, or specific scenarios with automatically generated lyrics
- Image-to-Music: Upload photos or videos to generate tracks that match the visual content's mood
- Enhanced Capabilities: Lyria 3 offers improved creative control over style, vocals, tempo, and musical complexity compared to previous models
- SynthID Integration: All generated tracks include imperceptible watermarks for AI content verification
- Multi-language Support: Available for users 18+ in 8 languages (English, German, Spanish, French, Hindi, Japanese, Korean, Portuguese)
- Copyright Protection: Built-in filters prevent mimicking specific artists and check outputs against existing content
- Platform Integration: Also available on YouTube's Dream Track for Shorts creators