Google Launches Lyria 3: Create Complete Songs with AI in the Gemini App

Google has officially launched Lyria 3, the latest version of its AI music model, integrating it directly into the Gemini app. Users over 18 can now create complete songs – with vocals, instruments, and full arrangements – directly from their phone or browser.

What is Lyria 3?

Lyria is Google's dedicated AI model for music generation. Since the first version launched in 2023, Google has worked closely with the music industry and experimented through projects like Music AI Sandbox. Lyria 3 is the result of this work – a model that understands music at a deeper level than its predecessors.

The new model can generate:

• Complete songs with vocals and lyrics
• Instrumental compositions across genres
• Music adapted to the mood of a video or image
• Variations based on natural language descriptions

Available in the Gemini App

The feature has been in beta since February 2026 and is now available to users over 18 in multiple languages. The goal is to democratize music creation – no musical background required.

Users can describe the type of music they want in plain text, and Lyria 3 generates a complete song. It's also possible to use existing images or videos as inspiration for the music.

Implications for CIOs

For businesses, this opens new opportunities in marketing, product presentation, and internal communications. AI-generated music can significantly reduce content production costs – from advertising jingles to background music for training videos and product demos.

Copyright and licensing questions around AI-generated music remain unresolved in many jurisdictions, but Google has been actively working with the music industry to address these issues.

Lyria 3 represents an important step toward a future where AI tools handle the entire media production chain – from text and images to audio and music.