Google DeepMind Lyria 3

Lyria 3 is the latest AI music generation model from Google DeepMind, integrated directly into the Gemini app. Launched in February 2026, it represents a significant leap from its predecessors by offering multimodal capabilities, allowing users to generate 30-second music clips using text, images, or video context. The model automates complex musical tasks such as lyric writing and song structuring (including intros, verses, and choruses). It also incorporates SynthID watermarking for audio provenance and safety. While Lyria 3 is the primary consumer-facing tool in Gemini, the Lyria 2 model remains the enterprise standard for developers via Vertex AI API.