Google Brings Music Generation to Gemini

Google has rolled out a new AI-powered music generation feature in its Gemini, expanding the platform’s creative capabilities. The tool uses DeepMind’s latest Lyria 3 to produce short, original tracks with lyrics, based on user prompts. Users can specify style, tempo, vocals, and mood, or upload photos and videos to guide the emotional tone of the music.

Lyria 3 offers more complex and realistic compositions than previous versions, giving users fine control over musical elements. While it cannot replicate specific artists directly, prompts referencing artists are interpreted as stylistic inspiration, not duplication.

Alongside Gemini, the AI model is integrated into YouTube through the Dream Track feature, now available globally, allowing creators to generate music for videos. All AI-generated tracks include a SynthID watermark, enabling identification and verification within Gemini. Users can also check whether uploaded tracks were AI-generated.

The feature is available worldwide for users 18+, supporting multiple languages including English, Spanish, French, German, Hindi, Japanese, Korean, and Portuguese. The rollout reflects Google’s effort to offer original music creation tools while providing transparency and safeguards for AI-generated content amid ongoing debates in the music industry.