Google adds music generation capabilities to the Gemini app

Source: TechCrunch

Google announced on Wednesday that it’s adding a music generation feature to the Gemini app, powered by DeepMind’s Lyria 3 music generation mode — currently in beta.

How the feature works

• Describe the song you want to create, and Gemini will generate a track along with lyrics.
  Example: “Create a comical R&B slow jam about a sock finding its match.” The app produces a 30‑second track and cover art generated by Nano Banana.
• You can also upload a photo or video, and the AI will compose a song that matches the mood of the media file.

Lyria 3 improvements

Lyria 3 builds on the previous generation of models, delivering more realistic and complex music tracks. Users can control elements such as style, vocals, and tempo. The model includes filters to prevent direct mimicking of existing artists; mentioning an artist’s name serves only as broad creative inspiration.

“Music generation with Lyria 3 is designed for original expression, not for mimicking existing artists. If your prompt names a specific artist, Gemini will take this as broad creative inspiration and create a track that shares a similar style or mood. We also have filters in place to check outputs against existing content.” – Google blog post

Availability

• Gemini app: Rolling out to all users aged 18+ worldwide, supporting English, German, Spanish, French, Hindi, Japanese, Korean, and Portuguese.
• YouTube Dream Track: The Lyria 3 model is also being made available to YouTube creators via the Dream Track feature, which was previously limited to U.S. creators. This release expands Dream Track globally.

Content identification

All songs generated with Lyria 3 receive a SynthID watermark to identify AI‑generated content. Gemini can analyze uploaded tracks and indicate whether they were created by AI, using the same SynthID technology.

Industry context

AI‑generated music has sparked mixed reactions:

• Platforms like YouTube and Spotify are integrating AI tools and signing deals with music labels to monetize AI‑generated tracks.
• The music industry is filing lawsuits over copyright concerns related to training data.
• Services such as Deezer have introduced tools to label AI‑generated music and curb fraudulent streams.

References

• DeepMind Lyria 3 model:
• Previous Lyria models:
• YouTube Dream Track support:
• SynthID detection in Gemini: