Lyria 3.5 music generation opens to everyone in Gemini app and API
Google has extended its Lyria 3.5 music generation model to the Gemini app and the Gemini API, opening AI songwriting to users worldwide instead of confining it to niche tools. The rollout on September 4 delivers what Google describes as its best-sounding music model to date, built around more expressive vocals and richer musical arrangements.
Inside the Gemini app, creation now starts with a genre selection or a plain-language description, a vocal or instrumental toggle, and templates aimed at concrete jobs such as background music or custom birthday tracks. Users can generate short clips or longer pieces, and the feature is live for everyone on the web and in the mobile app rather than gated behind an access list.
Lyria 3.5 music generation goes full-length in the API
Developers get the sharper step forward. The Gemini API now exposes a lyria-3.5-pro-preview model purpose-built for complete songs with coherent structure spanning verses, choruses, and bridges, plus fine-grained control over duration and song structure. It accepts text and image prompts and returns high-fidelity 44.1 kHz stereo audio with more natural vocals. Google AI Studio, Flow Music, and Google Vids run the same engine, so artists, video teams, and application builders draw on one generation backbone.
The update completes a fast climb inside Google's own Lyria family. Lyria 3 arrived in two shapes, a Clip model for 30-second vocal segments and a Pro preview aimed at longer pieces, while Lyria 3.5 extends that range across the full line, from short clips to full songs, under one model family. Lyria 3.5 music generation removes that earlier trade-off, letting one interface handle both a short hook and a multi-part composition.
The practical effect lands on production budgets. Teams that automate music for ads, video, or internal creative work previously juggled separate clip and long-form tiers, then stitched short outputs into longer cuts. Duration and structure controls in the API remove that assembly step, and Google's prompt guidance for Lyria models describes steering a track with up to ten reference images or PDFs to set its emotional baseline, plus negative prompts to exclude unwanted elements.
What the rollout signals
The distribution decision carries as much weight as the model itself. Google is shipping one generation engine to consumers, artists, and developers on the same day, and template-driven entry points in the app target non-musicians who would never open a digital audio workstation. For decision-makers weighing Lyria 3.5 music generation, the practical test is workflow fit: the same API surface used for text and image models now returns finished stereo tracks, so audio generation plugs into existing pipelines instead of requiring a separate stack.
Why this matters
Music generation has crossed from 30-second clips to complete, structurally coherent songs, and Google has made that capability a mainstream product surface inside Gemini rather than a research demo. Teams building on Google's stack can treat full-length generated audio as a shipping feature now, which raises the bar for what an AI music tool should deliver from day one.
Sources
Create your best tracks yet with Lyria 3.5 in Gemini.
Generate music with Lyria 3.5 | Gemini API | Google AI for Developers
Lyria 3.5 Pro preview | Gemini API - Google AI for Developers
Generate music with Lyria | Gemini Enterprise Agent Platform | Google Cloud Documentation
Release notes | Gemini API | Google AI for Developers
Lyria 3 | Gemini Enterprise Agent Platform | Google Cloud Documentation
Lyria | AI Music Generator | Gemini Enterprise Agent Platform | Google Cloud Documentation
Google models | Gemini Enterprise Agent Platform | Google Cloud Documentation
Models | Gemini API | Google AI for Developers
Lyria music generation prompt guide | Gemini Enterprise Agent Platform | Google Cloud Documentation
system_prompts_leaks/Google/gemini-3.1-pro.md at main · asgeirtj/system_prompts_leaks
Related Articles
- Google Expands Gemini Ecosystem with Native macOS App and Lyria Music Tools
- Google Launches Gemini 3.1 Flash TTS for AI Audio
- Stability AI Debuts Stable Audio 3.0 with Six-Minute Music Generation
✔Human Verified
Researched and cross-referenced against primary sources by the Bytevyte editorial team. This article was generated with the assistance of artificial intelligence and reviewed by the Bytevyte editorial team.