Google is turning Gemini into a mini music studio, giving users a new way to create songs from simple prompts. With the addition of Lyria 3, an AI music model from Google DeepMind, Gemini can now generate short, complete tracks with vocals and even create matching album cover art. Like many early AI video tools, each song is currently capped at 30 seconds, keeping outputs quick, shareable, and easy to experiment with.
The feature is launching in beta and will be available to users aged 18 and up. Google says it’s rolling out gradually across multiple languages, including English, German, Spanish, French, and Japanese. There’s no separate registration process—eligible accounts will see access enabled automatically as the rollout expands.
Creating a song works much like chatting with any AI assistant: you describe what you want, and the model produces a result. You can specify the genre, mood, tempo, or a detailed creative idea. Google even suggests prompt styles like “a nostalgic Afrobeat song for my mother,” which gives the AI clear direction for tone and theme. For people who want the music to match existing content, Gemini also supports uploading images or videos, letting Lyria 3 generate audio that fits the vibe of what you’ve shared.
Gemini can also package the output into a more complete “release-style” bundle. Alongside the music, it can produce a fitting cover image using Google’s Nano Banana image model, based on your text prompt or uploaded media. The finished result includes the track, cover art, and a title, and can be downloaded or shared through a share link generated by the app.
To help with transparency around AI-generated media, all songs created through this system include SynthID, an invisible watermark that identifies the content as AI-generated.
What makes Lyria 3 different is how far Google says it has pushed realism and musical complexity compared with earlier versions. Lyria has been tested since 2023 through limited experiments and internal trials, but this marks a more official step where music generation becomes a standard feature available to everyday Gemini users. Lyria 3 is designed to offer more control over style, tempo, and vocals, and it can also generate lyrics automatically from the user’s prompt. Google frames the purpose as creative self-expression rather than perfectly polished, radio-ready hits—more like a fast, flexible songwriting companion than a replacement for professional production.
Google is also expanding where this technology shows up. Lyria 3 is set to integrate with YouTube Dream Track, allowing creators to generate custom soundtracks for Shorts. After initially being limited to the United States, the feature is now rolling out more broadly to users worldwide, making it easier for creators to produce quick, personalized music that fits short-form videos.
With Gemini gaining AI song generation, lyric creation, and album art in one flow, Google is positioning its assistant as more than a chatbot. It’s becoming an all-in-one creative tool for anyone who wants to experiment with music ideas, draft a soundtrack for a video, or turn a concept into a shareable track in minutes.
