Krux
February 21, 2026
Google's Lyria 3 Turns Text Into Music in Gemini
Published: February 21, 2026 at 3:44 AM
Updated: February 21, 2026 at 3:44 AM
100-word summary
Google DeepMind just dropped Lyria 3 in beta, letting Gemini users create 30-second tracks from text, photos, or videos with auto-generated lyrics and cover art. Think of it as Instagram for your ears: describe a vibe or upload a sunset pic, and get a custom soundtrack perfect for TikTok or Reels. The AI handles genre, mood, tempo, and vocals while blocking artist copycats. Each track gets watermarked via SynthID so you can spot AI audio. Available now on desktop (mobile coming soon) for users 18+ in eight languages, with YouTube Dream Track integration expanding creator access. As AI music tools proliferate, expect short-form content to become hyper-personalized sonic experiences.
What happened
Google DeepMind just dropped Lyria 3 in beta, letting Gemini users create 30-second tracks from text, photos, or videos with auto-generated lyrics and cover art. Think of it as Instagram for your ears: describe a vibe or upload a sunset pic, and get a custom soundtrack perfect for TikTok or Reels. The AI handles genre, mood, tempo, and vocals while blocking artist copycats. Each track gets watermarked via SynthID so you can spot AI audio. Available now on desktop (mobile coming soon) for users 18+ in eight languages, with YouTube Dream Track integration expanding creator access.
Why it matters
As AI music tools proliferate, expect short-form content to become hyper-personalized sonic experiences.