Google Rolls Out Lyria 3 Pro With Three-Minute AI Music Tracks
Google's Lyria 3 Pro generates structured songs up to three minutes long, rolling out across Vertex AI, Gemini, Vids and ProducerAI with SynthID watermarking.

Updated
Why it matters
- Lyria 3 Pro generates tracks up to 3 minutes long with control over intros, verses, choruses and bridges
- Available in public preview on Vertex AI, plus AI Studio, the Gemini API, Google Vids, the Gemini app and ProducerAI
- All outputs carry SynthID watermarking; Google says the model does not mimic named artists
Google has launched Lyria 3 Pro, a music generation model that produces tracks up to three minutes long with structural control over intros, verses, choruses and bridges. The release, announced roughly a month after Lyria 3 debuted, pushes Google's most advanced music model into Vertex AI, Google AI Studio, the Gemini API, Google Vids, the Gemini app and the collaborative tool ProducerAI.
The upgrade matters because it moves AI music generation from short novelty clips toward production-grade output. Song-length tracks with recognizable structure address a core limitation of earlier models, which struggled to maintain coherence beyond brief snippets. For businesses and developers, that opens practical use cases: bespoke gaming soundtracks at scale, integrations into creative tools, and music and video platforms.
Lyria 3 Pro "better understands musical composition," according to Google, so users can prompt for specific song elements and generate material with complex transitions. The base Lyria 3 model brings "improved musical awareness and structural coherence" to developers building creative tools via the Gemini API.
Where it is available
- Vertex AI: Lyria 3 Pro is in public preview for businesses that need on-demand audio at scale.
- Google AI Studio and the Gemini API: Lyria 3 sits alongside Lyria RealTime in AI Studio for developers.
- Google Vids: Custom music in the AI-powered video app starts rolling out this week to Google Workspace customers and Google AI Pro and Ultra subscribers.
- Gemini app: Longer Lyria 3 Pro generations are available now for paid subscribers, aimed at vlogs, podcasts and tutorial videos.
- ProducerAI: The collaborative tool, built by musicians, uses Lyria 3 Pro to offer an "agentic experience" for artists, producers and songwriters. It is available globally to free and paid subscribers.
Industry partnerships and named artists
Google says it developed Lyria 3 through its Music AI Sandbox, a suite of experimental tools for musicians, producers and songwriters whose insights shaped the model. Grammy-winning producer Yung Spielburg used Lyria in composing and producing the score for the Google DeepMind short film "Dear Upstairs Neighbors." DJ and producer François K used Lyria iteratively to create a song Google says is set for release.
Guardrails and provenance
Google trained Lyria 3 on "materials that YouTube and Google has a right to use under our terms of service, partner agreements, and applicable law." The company states that Lyria 3 and Gemini do not mimic artists; if a prompt names a creator, the model treats that name as broad inspiration rather than an instruction to imitate. Google applies filters that check outputs against existing content, and users must follow Terms of Service and Gen AI prohibited use policies covering intellectual property and privacy rights.
Every output from Lyria 3 and Lyria 3 Pro carries SynthID, Google's imperceptible watermark for identifying AI-generated content.
The rollout puts Google in direct competition with Suno and Udio in the fast-growing AI music market, while its licensing posture and watermarking signal an attempt to preempt the copyright disputes now shaping generative AI policy. How widely creators adopt structured, three-minute AI tracks — and whether SynthID detection satisfies platforms and rights holders — will determine whether Lyria 3 Pro becomes a production tool or stays a novelty.
Original: blog.google
More from James Calloway
Show full bio
News editor covering industry trends and analytics at AI In Context.
119 articles
Related articles
- Google Launches Lyria 3.5 in Flow Music
- Google's Lyria 3 Brings AI Music Generation to the Gemini App
- Google Launches Gemini 3.1 Flash TTS With Audio Tags and 70+ Languages
- Google Ships Upgraded Gemini 2.5 Flash Native Audio and Live Translation
- Google Ships Gemini 3.8 TTS Models With Voice Cloning and Direction