Google Launches Lyria 3.5 Music Model in Gemini App as Suno Faces Copyright Ruling

Google Expands Lyria 3.5 Access Across Gemini Platforms

Google has launched Lyria 3.5 in the Gemini app and the Gemini API, making its advanced music generation model available to all Gemini users globally on web and mobile platforms1

1

2

2

. The model, which first launched in Google Flow Music on 29 July, now extends to Google AI Studio and Google Vids, giving developers and creators broader access to AI music generation capabilities3

3

. Users can now generate full songs with verses, choruses, and bridges at 44.1kHz stereo quality from text or image prompts1

1

. The expansion positions Google to compete in an increasingly crowded AI music market where platforms like Deezer report that more than half of new uploads are AI-generated, with around 90,000 fully AI tracks arriving daily1

1

.

Enhanced Features Enable Users to Create Your Best Tracks

Lyria 3.5 delivers significant improvements in audio quality and creative control for AI music generation. The model produces expressive vocals with more realistic and emotionally nuanced performances, alongside richer musical arrangements that showcase complex melodic structures2

2

3

3

. In the Gemini app, users can easily select or describe their genre and choose between vocal or instrumental styles, with new templates designed to jumpstart creativity for everything from background music to custom birthday tracks3

3

. The platform now offers flexibility to choose between short and longer tracks, with Lyria 3 Clip producing 30-second clips and Lyria 3.5 generating tracks lasting a couple of minutes1

1

2

2

. Developers working through the API can supply up to ten images, write custom lyrics using section tags, and use timestamps to specify when instruments enter, creating high-fidelity tracks for applications ranging from video backing tracks to brand jingles and personalized ringtones1

1

3

3

.

Copyright Safeguards Through SynthID Watermarking and Content Filters

Google implemented specific guardrails in Lyria 3.5 to address copyright concerns in AI music generation. Every track carries SynthID watermarking that Google describes as imperceptible, designed to identify AI-generated content1

1

. Safety filters actively block requests for particular artist voices and prevent the generation of copyrighted lyrics1

1

. These protections arrive as the AI music industry faces mounting legal scrutiny. On 31 July, a Munich Regional Court ruled against Suno for copyright infringement in the first judgment of its kind in Europe, finding that works were memorized and reproducibly contained in model versions1

1

. The court held Suno responsible for infringing outputs and rejected arguments that complex prompting shifted responsibility to users. Significantly, the Munich court asserted jurisdiction over training conducted entirely in the United States, ran its own analysis of US fair use, and concluded it did not apply1

1

. The injunctions cover training on copyrighted works, retaining them in model parameters, and offering trained models in Germany, with language reaching copying inside the United States1

1

.

Market Implications and the AI Music Saturation Challenge

Google enters a market already grappling with AI-generated content saturation. Deezer reports that AI tracks account for between 1% and 3% of actual listening, yet up to 85% of the streams they generated in 2025 were fraudulent1

1

. While Deezer has been actively fighting back against this flood, Spotify has not taken similar measures1

1

. The Munich ruling against Suno could reshape how AI music companies operate in Europe and potentially influence US practices. Watch for appeals of the judgment, which is not yet final, and whether other courts adopt similar reasoning about jurisdiction over AI training conducted outside their borders1

1

. Google's proactive approach with content filters and watermarking suggests the company is positioning Lyria 3.5 to navigate these legal challenges while capturing market share from competitors facing litigation. The success of this strategy depends on whether Google's safeguards prove sufficient to satisfy rights holders and regulators as AI music generation becomes mainstream across the Gemini app and API.