Google Releases Lyria 3 Pro with Full Song Structure Control

Google

Mar 25, 2026 · Updated Apr 25, 2026

Google released Lyria 3 Pro with structural awareness of song architecture — intros, verses, choruses, and bridges. Tracks now reach up to 3 minutes, enabling complete song generation across the Gemini app, Vertex AI, and Google Vids.

Lyria 3 Pro is Google's latest AI music generation model, expanding on Lyria 3 with structural awareness of how songs are built. The model understands musical architecture — intros, verses, choruses, and bridges — and generates tracks with complex transitions. Track length has increased from 30 seconds to up to 3 minutes.

According to Google, what makes this upgrade significant is specifically the architectural understanding — not just the longer duration. The model's awareness of song structure is what enables complex transitions and section-based composition in a way that shorter clip models cannot.

Lyria 3 Pro is available now via the Gemini API, Google AI Studio, the Gemini app for Google AI Pro and Ultra subscribers, Vertex AI, Google Vids for Workspace customers, and ProducerAI. If you're building audio features or working with AI-generated soundtracks, full song territory is now within reach.

View the full update on blog.google

Google AI

@GoogleAIMar 25

Last month, we released Lyria 3, enabling you to create tracks with lyrics from text, image, or video prompts. Now, we’re introducing Lyria 3 Pro, which expands upon our music generation model to offer additional advanced capabilities. What’s really special about this upgrade is that the model now understands the architecture of music. This makes it possible to prompt for intros, verses, choruses and bridges + generate songs with more complex transitions. You can also create tracks up to 3 minutes long, a big change from previous models that were limited to 30 second tracks. Use Lyria 3 Pro to build upon your existing creativity. We’re excited for your beats to drop 🎶

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

See all AI news & updates from Google →

Keep reading

Google Releases Lyria 3 Music Generation Models via Gemini API

Google DeepMind's Lyria 3 music generation models are now in public preview via the Gemini API and Google AI Studio. Lyria 3 Pro handles full-length tracks; Lyria 3 Clip targets rapid generation — every output carries a SynthID watermark.

Google DeepMindFeb 18

Lyria 3 Brings AI Music Generation with Lyrics to the Gemini App

Google DeepMind's Lyria 3 is live in the Gemini app - describe a mood or upload a photo and get a 30-second track with AI-generated lyrics and vocals. It's Lyria's first wide public release after two years of controlled experiments.

Google AI StudioApr 1

Google AI Studio Launches Music Playground to Prototype and Export Audio Code

Google AI Studio integrated the Lyria 3 music model family into a new interactive playground featuring a Pro preview for three-minute tracks. Users can now build precise musical prompts and export the results directly into functional code for the Gemini API. This update transitions high-fidelity audio generation from a standalone creative tool into a programmable developer utility.

Google GemmaJun 4

Google Magenta RealTime 2 Turns MacBooks into Live AI Music Instruments

Google's Magenta Project released Magenta RealTime 2 (MRT2), an open-weight, open-source live music model. It enables low-latency, real-time music synthesis natively on Apple Silicon MacBooks using MIDI, text, and audio inputs. This allows musicians to play AI-generated music as an instrument directly on their device, fostering new creative workflows.