Google launches Lyria 3 Pro music generator, claims training data is rights-cleared
Google has released Lyria 3 Pro, its latest AI music generation model capable of creating tracks up to three minutes long with improved understanding of musical structure. The model is available through Gemini, Google Vids, Vertex AI, and Google AI Studio. Google claims the training data comes from sources it has contractual and legal rights to use.
Google Launches Lyria 3 Pro Music Generator
Google has released Lyria 3 Pro, an upgraded version of its AI music generation model that can produce tracks up to three minutes in length. The model represents an incremental improvement over Lyria 3, which launched in February 2026.
Key Capabilities
Lyria 3 Pro demonstrates improved structural understanding compared to its predecessor, handling musical elements like intros, verses, choruses, and bridges with greater accuracy. According to Google, the model does not imitate specific artists when their names appear in prompts, but instead uses them as creative inspiration.
The model is integrated across multiple Google products: Gemini (for paying subscribers), Google Vids (for Workspace customers), Vertex AI (for enterprise customers), and Google AI Studio (for developers). ProducerAI, Google's collaborative music generation tool, also leverages Lyria 3 Pro.
Training Data and Rights Claims
Google states the model was trained exclusively on materials "that YouTube and Google has a right to use under our terms of service, partner agreements, and applicable law." The company declined to provide additional details about training data composition, sources, or licensing agreements.
This claim comes amid intensified scrutiny of AI music generator training practices. Suno, the primary competitor in high-quality AI music generation, currently faces multiple lawsuits from record labels alleging copyright infringement. The legal battles center on whether generative AI companies obtained proper licensing or rights to the music used for training.
Copyright Protection
All content generated by Lyria 3 Pro includes an invisible SynthID watermark, Google's technique for marking AI-generated audio. This watermark persists through common audio processing operations and can be detected to verify synthetic origin.
Market Context
Lyria 3 Pro enters a rapidly developing but legally contested market. Suno remains the only comparable high-quality competitor, making this one of the few segments where consumer-facing AI music generation is commercially available. The relative scarcity of viable options reflects both technical difficulty and legal risk in the space.
What This Means
Google is positioning itself as the rights-conscious alternative to Suno by explicitly claiming contractual legitimacy for its training data. However, the company's refusal to disclose specifics makes independent verification impossible. For enterprise and developer customers, availability on Vertex AI and Google AI Studio expands access to music generation capabilities. The invisible watermarking adds a layer of AI-generated content attribution, though downstream detection reliability remains untested at scale.
Related Articles
Google launches Gemini 4 Hydrogen: 1M-token context, $1.50 input and $7.50 output per 1M tokens
Google announced Gemini 4 Hydrogen on November 12, 2026, with native 1 million-token input and up to 1 million output tokens. It costs $1.50 per million input tokens and $7.50 per million output tokens, and Google claims it scores 51 on the Artificial Analysis Intelligence Index at its default high thinking level.
Translated releases Lara Mini, a 7.2B-parameter MoE research model with 600M active parameters
Translated announced Lara Mini on November 12, 2026, a research model trained from scratch on 6 trillion tokens across English and 44 other languages, including all 24 official EU languages. The mixture-of-experts system has 7.2 billion parameters and activates 600 million per token.
inclusionAI releases Duo 1.0 causal-base, a full-duplex multimodal world model initialized from MiniMax H3
inclusionAI has published the Duo 1.0 causal-base weights on Hugging Face. Duo is described as a full-duplex multimodal world model for real-time audio-visual interaction, initialized from MiniMax H3 and trained further by the Duo team.
Endless Frontier open-sources Fx-Bio-1031, a 125B MoE bioinformatics model with 512K context
Endless Frontier (无界前沿) has open-sourced Fx-Bio-1031, a bioinformatics model with 125B total parameters and 6B active per token, with a 512K context window. The company claims 70.59% Pass@5 on the Difficult subset of Anthropic's BioMysteryBench, against 47.06% for GPT-6-Astra and 52.94% for Claude Opus-5.
Comments
Loading...