DeepMind Releases Lyria 2 for AI Music Creation

Google DeepMind is back with Lyria 2, a new version of its AI music model and it’s starting to feel like music production from the future. Part of the Music AI Sandbox, Lyria 2 blends creativity with high-tech control in a way that feels like flying a spaceship.

With 48kHz stereo output, Lyria 2 creates music in high audio quality, ready for pro use. But it’s not just the sound, it’s the experience. Musicians can type in a description like “jazzy synth with dreamy vocals” and it will generate exactly that.

It has three main tools that make up the core:

1. Create – Start a new track from a simple idea.
2. Extend – Add new parts to what you’ve already got.
3. Edit – Change specific parts of a song without touching the rest.

The system even features a RealTime mode, where Lyria responds on the go. It’s like having a co-pilot in your creative journey. For musicians who’ve dreamed of instant, intuitive tools, this gets close.

To keep things honest, DeepMind adds a special watermark using SynthID, tagging all AI-made audio without messing with the sound. It’s a quiet step to keep track of what’s human and what’s not.

For now, Lyria 2 is in limited testing with U.S.-based creators, but DeepMind says it’s listening closely to artist feedback. The aim isn’t to replace musicians it’s to push creativity.

More info on Lyria 2 on Google’s Deepmind.

For more such articles, read :

• Ableton Live 12.2 Is Here: Workflow Upgrades, New Devices & More
• Why AI Won’t Kill Creativity, But Will Demand More of It