Google Launches AI Music Model Lyria 3 in Gemini App
- Google’s Lyria 3 model generates full songs from text prompts or images, available in the Gemini app.
- The model produces coherent tracks but is limited to a maximum length of 30 seconds.
- Rivals like Suno and Udio offer longer songs with more controls and mature workflows.
- Lyria’s genre range includes pop, afrobeat, R&B, and light hip-hop but struggles with niche genres.
- All tracks generated include SynthID, Google’s AI watermark for transparency and safety.
Lyria’s release marks Google’s entry into AI music generation for public use via the Gemini app, offering users over 18 years old the ability to create short tracks from descriptions or images. Despite its innovative approach, it faces competition from existing models like Suno and Udio that provide longer compositions with advanced features.
While Lyria provides a fun tool for generating quick musical ideas, its limitations in track length and genre flexibility position it differently compared to its rivals’ capabilities in producing fully structured songs. (Source)
