Google adds music-generation capabilities to the Gemini app

Google announced on Wednesday that it's adding a music-generation feature to the Gemini app. The company is tapping DeepMind's Lyria 3 music-generation model to power the new capability, which is currently still in beta.

To use the feature, you describe the kind of song you want to create, and the app generates a track with lyrics. For example, you could ask Gemini to make a "comical R&B slow jam about a sock finding its match," and the app will produce a 30-second track along with cover art created by Nano Banana.

Google also said you can upload a photo or video, and the AI-powered tool will generate a song that matches the media's mood.

The company noted that Lyria 3 is an improvement over previous generations of its music models, producing tracks that are more realistic and more complex. Users will also be able to adjust and control elements such as style, vocals, and tempo.

Alongside bringing Lyria 3 into the Gemini app, Google is also making the model available to YouTube creators through Dream Track on YouTube. This feature helps creators generate AI-made music tracks. Dream Track had previously only been available to YouTube creators in the U.S., but with this update, Google is expanding the feature globally.

Google said users won't be able to mimic an artist directly, but if an artist's name is included in a prompt, Gemini may generate a track that matches a similar style or mood. The company also acknowledged questions about whether this could make it easier for others to decode or replicate a particular artist's musical style.

"Music generation with Lyria 3 is designed for original expression, not for mimicking existing artists. I suppose your prompt names a specific artist; Gemini will treat it as broad creative inspiration and create a track with a similar style or mood. We also have filters in place to check outputs against existing content," the company said in a blog post.

Google added that all songs generated with Lyria 3 will include a SynthID watermark to identify AI-generated content. The company said it's also adding tools in Gemini to help detect AI-generated music using SynthID. Users will be able to upload tracks and ask Gemini whether the audio is AI-generated.

Music generation is rolling out to Gemini users aged 18 and up worldwide, with support for English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese.

AI-generated music continues to spark mixed reactions among artists and listeners. On the one hand, companies like YouTube and Spotify are adopting AI tools and signing agreements with music labels to monetise AI-generated music. On the other hand, AI model and tooling companies are facing lawsuits from the music industry over copyright concerns related to their training data. Platforms such as Deezer have also introduced tools to label AI-generated music, in an effort to reduce fraudulent streaming activity tied to this content.