You can now use Gemini to turn prompts and photos into music: How it works

Google adds Lyria 3-powered music creation to Gemini, letting users generate 30-second songs from text, photos or videos, with AI watermarking and Dream Track support

Google is introducing a music-generation feature to the Gemini app that lets users create songs using text prompts, photos or videos. Powered by Google’s Lyria 3 model, the feature is currently available in beta. According to Google, users can describe the type of song they want, the mood or theme they have in mind, and Gemini will generate a 30-second track. The app also produces matching cover art using Nano Banana. These tracks can be downloaded or shared through a link. Google said that the feature is designed for quick, creative expression rather than professional music production.

Google’s music generation feature: How it works

According to Google, Lyria 3 builds on Google’s earlier music models with a few key changes. The company said that users don’t need to write their own lyrics — the model can generate them automatically based on the prompt. Additionally, it offers more control over details like style, vocals and tempo. As mentioned, the Gemini app creates 30-second songs along with custom cover art generated by Nano Banana.  

Google said that there are two main ways to use the feature:

• Text to track: Users can type a description of the kind of song they want, including a genre, mood, memory or even an inside joke. 
• From photos or videos to track: Users can also upload a photo or video. Gemini will analyse the content and generate a song with lyrics that match the mood. 

For YouTube’s Dream Track

In the blog, Google stated that creators can also try Lyria 3 through YouTube’s Dream Track feature. It is currently available in the US and is gradually rolling out to creators in other countries. For reference, Dream Track is a suite of experimental tools that enable creators to generate unique music for their Shorts. It creates an AI-generated soundtrack based on users’ text prompts. 

According to Google, YouTube Shorts creators will be able to generate higher-quality soundtracks for their videos with Lyria 3. The company said creators can produce short songs with lyrics or background music that aligns with the mood of their content.

 

New audio verification capabilities

Google said that all tracks created in the Gemini app will include SynthID, a digital watermark designed to identify content generated by its AI. The watermark is not noticeable to users. Google is extending its Gemini verification feature to check audio files, in addition to photos and videos. Users can upload a file and ask whether it was made using Google AI. Gemini will then check for the SynthID watermark and analyse the content before giving a response.

AI Music Safeguards

Google said that Lyria 3 is built to support original music creation and is not meant to copy specific artists. If a user mentions an artist in a prompt, Gemini will treat it as general inspiration and generate a track with a similar style or mood, rather than imitating that artist. The company added that it has put filters in place to compare outputs against existing content. In addition, users can report content that they believe violates their rights or someone else’s rights.  

Availability 

Lyria 3 is now available in the Gemini app for users aged 18 and above in English, German, Spanish, French, Hindi, Japanese, Korean and Portuguese. Google said it plans to add support for more languages over time. As per Google, the feature is rolling out on desktop starting today and will reach the mobile app in the coming days. Google AI Plus, Pro and Ultra subscribers will have access to higher usage limits.