Google has introduced Lyria 3, its latest and most capable generative music model, now integrated directly into the Gemini AI app — a significant step forward in AI-driven music creation.
Unveiled in mid-February 2026, Lyria 3 represents the next generation of AI music technology developed by Google DeepMind. It allows users to generate original 30-second music tracks complete with melodies, instrumentation, and even lyrics using simple prompts — either text descriptions or uploaded images and videos.
How It Works
With Lyria 3 now available in beta within the Gemini app, users aged 18 and older can create tracks by:
- Typing a text description of a song idea (e.g., mood, genre, tempo).
- Uploading a photo or video to inspire music that matches the visual vibe.
- Receiving a 30-second audio clip generated by AI with custom instrumentation and vocals.
Each piece of music also comes with custom AI-generated cover art, making it easy to share and enjoy. Google is positioning this as a tool for creative expression rather than professional music production, suitable for social posts, personal projects, and lighthearted audio experiments.
Personalization and Creativity
Lyria 3 offers users some control over the style and output of their AI-generated tracks — including genre selection, vocal presence, and tempo variation. This improves on previous versions by delivering more cohesive and musically complex outputs, tailored to the specifics of the prompt.
Beyond simple text input, the ability to upload images and video adds a multimodal creative twist — meaning your photos or memories can directly inspire original music.
Responsible Use and Watermarking
To address industry concerns around AI-generated media and copyright, Google embeds each track with SynthID, its invisible watermarking technology designed to identify AI-generated audio content. Users can even upload files to check if they were created using Google’s AI tools.
Google also states that while Lyria 3 can take stylistic cues from named artists, it is not designed to mimic or recreate existing artists’ actual sounds, focusing instead on original creation within broad stylistic boundaries.
Availability
Lyria 3 is rolling out first in the Gemini app on desktop, with mobile support arriving soon. It’s available globally in multiple languages, including English, Spanish, French, Hindi, German, Japanese, Korean, and Portuguese.
What This Means
Google’s launch of Lyria 3 brings AI-powered music creation into the mainstream for everyday users — from social media creators to casual music fans. While the tool is not marketed as a full-blown studio replacement, its accessibility and ease of use could reshape how people think about composing and sharing music in the digital era.
