Gemini users can make 30-second songs at no extra cost, while developers pay four cents a clip with no free API tier.

Kenning's take

What it does

Lyria 3 is a music generation model from Google DeepMind. Compared with earlier Lyria models it writes its own lyrics, gives more control over style, vocals and tempo, and produces more complex tracks. It launched in beta in the Gemini app, where it makes 30-second tracks with cover art, and in YouTube Shorts Dream Track; developers can call it in the Gemini API, where Google now lists it as a legacy model next to Lyria 3.5.

• Text, photo or video to a 30-second track
• Lyrics written for the user, or instrumental tracks
• Control over style, vocals and tempo
• Cover art made by Nano Banana
• SynthID watermark on every track

Who it suits. Gemini app users who want a quick custom song, and developers adding short music clips to apps.

Key facts

Category

AI models

Part of

Google Gemini

Made by

Google DeepMind

Based in

United States

Released

18 February 2026

How to get it

in the Gemini app for users aged 18 and over (higher limits on Google AI Plus, Pro and Ultra), in YouTube Shorts Dream Track, and in the Gemini API as lyria-3-clip-preview

Works on

Web, iOS, Android, API

In the UK

Yes

Free plan

Available to Gemini app users aged 18+; no free tier in the API

API price

$0.04 per song for Lyria 3 Clip Preview (30s); $0.08 per song for Lyria 3 Pro Preview (full song)

FAQ

Facts from the maker's own pages and the sources listed. Checked 2026-10-11.