Google DeepMind just dropped Lyria 3.5, its latest AI music generation model, rolling it out directly inside Google Flow Music. The upgrade touches every layer of the stack: how well the model follows your creative directions, how expressive the singing sounds, how naturally arrangements move between notes, and how precisely you can dial in tempo and song length.
From research toy to production tool
The Lyria family has moved fast. Lyria 3 launched and let users generate 30-second high-fidelity music tracks from text or image prompts through the Gemini app. Lyria 3 Pro followed, expanding to full songs up to three minutes long, and introduced structural coherence so tracks could include proper intros, verses, choruses, and bridges instead of sounding like a random assemblage of musical ideas. Now Lyria 3.5 pushes the model further, focusing on quality rather than length.
Google Flow Music itself is a comprehensive AI-driven platform designed to let users compose, publish, and share original music, featuring a "Chat with Producer" interface for creating full-length songs with dynamic vocals and rich musicality. The platform also draws on Google's Gemini, Veo, and Nano Banana models for chat, video, and images. Think of it less as a Suno clone and more as an AI-native DAW (Digital Audio Workstation) being assembled piece by piece.
What actually changed in 3.5
The four pillars of the Lyria 3.5 upgrade, straight from the official announcement:
- Improved musicality: Richer, more complex melodic structures that sound more natural.