Google has released Lyria 3, a high-fidelity AI music generator built directly into the Gemini app. According to Google's music generation overview, the tool transforms text or image prompts into 30-second tracks complete with instrumentals, vocals, and lyrics. All generated tracks are watermarked with SynthID, Google's imperceptible AI content identifier, and users can now verify whether audio files were created using Google's AI.
Two Ways to Generate: Text and Visual Input
The standout feature of Lyria 3 is its dual input capability. Users can generate music from:
Text prompts: Describe a mood, genre, or specific sound ("80s synth-pop with upbeat, danceable rhythm and airy female vocals") and Gemini generates a matching 30-second track
Image or video input: Pair a photo, slide, or video with a prompt, and Gemini generates lyrics and music that match the visual moment
This image-to-soundtrack workflow is particularly useful for content creators working with existing visual assets—slides, storyboards, or reference footage—who need matching audio without describing the scene in text.
Building Tracks: Templates and Customization
For users who prefer starting points over blank canvas creation, Lyria 3 includes:
Template gallery: Pre-built starting points that users can customize and refine
Custom cover art: Generated through Nano Banana, Google's image generation model, allowing users to create matching visual assets for their tracks
SynthID Watermarking and Content Verification
Every track generated in Gemini is embedded with SynthID, Google's imperceptible watermark for identifying AI-generated content. Users can now upload any audio file and ask Gemini whether it was generated using Google AI. Gemini checks for the SynthID watermark and applies its own reasoning to return a response. This verification capability addresses growing industry concerns about AI-generated content attribution and authenticity.
Lyria 3 vs. Suno: Where It Sits in the Landscape
Lyria 3 enters a competitive space dominated by tools like Suno, which also generates full songs with vocals from text prompts. The key differences:
Integration: Lyria 3 is built into Gemini (no separate app required), while Suno is a dedicated music-creation platform
Input types: Lyria 3 accepts image and video prompts; Suno focuses on text-based generation
Track length: Lyria 3 generates 30-second tracks; Suno produces longer, full-length songs
Provenance: Lyria 3 includes SynthID watermarking for content verification; Suno does not offer this feature
For creators already using Gemini, Lyria 3 offers quick custom music generation without switching tools. For those seeking longer compositions or a dedicated music-creation interface, Suno remains the alternative.
Availability and Age Requirements
Lyria 3 is available in all countries where the Gemini app is available. Access requires users to be 18 or older. The tool is free to use within the Gemini app, with no separate subscription or per-track fees.
