Google's Lyria 3.5 Music Model Rolls Out to All Gemini Users
Google announced on 09-05 that Lyria 3.5, its most advanced music generation model, is now available to all Gemini users worldwide on both web and app, and also live on AI Studio and the Gemini API. The company calls it its best-sounding, most expressive music model, and early user feedback has been strikingly positive.
Confirmed
- Lyria 3.5 is available in the Gemini app, on the web, in AI Studio, and via the Gemini API
- The model delivers richer arrangements, more expressive vocals, and higher fidelity
- Users can pick a genre, choose vocal or instrumental versions, use new templates, and generate shorter or longer complete tracks
- Users start by selecting "Create music" in the tools menu and describing the desired style
- The Google GenMedia team shipped developer documentation alongside the release; commenter AIAndrew highlighted vocal realism with genuine emotional weight, dynamic breathing, and phrasing as the biggest gains
- User DeryaTR reported generating a complete track in 30 seconds with stunning results, calling it "epic" amid a busy week of model releases
Why it matters
- This is a major upgrade to Google's music generation, offered to all Gemini users globally rather than as a limited test, lowering the barrier for everyday AI music creation
- Simultaneous AI Studio and Gemini API access lets developers integrate the model into third-party apps, putting it in direct competition with products like Suno
2026-09-05 ~ 2026-09-06 · 9 related posts
Primary sources
- [source] Google ships Lyria 3.5 music model in Gemini with richer vocals and arrangements — GeminiApp · 2026-09-05
- [source] Google rolls out Lyria 3.5 music generation to all users globally — GeminiApp · 2026-09-05
- [source] Google launches Lyria 3.5 music model with human-like vocals, now in AI Studio — AI_Andrew · 2026-09-06
6 near-duplicate retellings: GeminiApp · GeminiApp · GeminiApp · cedric_chee · gaganghotra_ · DeryaTR_