Google released Lyria 3.5, its latest music generation model, into Google Flow Music on July 29, 2026. The update sharpens the tool for songwriters, producers, and other creatives by delivering better musicality, lyrics, vocal quality, and finer creative controls.
"We're launching Lyria 3.5 in Google Flow Music, with advances across musicality, lyrics, vocals, and creative control," Google said in a company blog post.
The model's launch highlights the growing role of AI for Creatives in music production, giving artists practical tools for songwriting rather than gimmicks. Flow Music already lets users generate original tracks from text prompts, but Lyria 3.5 brings a notable leap in output quality.
More Natural Melodies and Better Lyrics
Lyria 3.5 creates richer, more complex melodic structures that sound less synthetic. The engine understands musical phrasing more naturally, producing chord progressions and motifs that feel intentional rather than randomly assembled.
Lyrics generation has also improved. The model follows prompts more closely and shows a stronger grasp of song structure. It can maintain coherent verses, choruses, and bridges, helping users craft songs that hold together across multiple sections.
Expressive Vocals and Deeper Creative Control
The vocal synthesis in Lyria 3.5 adds emotional nuance and more realistic pronunciation. Voices sound less flat, with subtle inflections that make lead lines or backing harmonies feel more human. For vocalists and songwriters, the new vocal quality is a step forward, and those following an AI Learning Path for Vocal Artists & Songwriters can test the capabilities directly.
On the production side, users now get more direct control over tempo and the duration of generated audio. Instead of guessing how long a track will be, creators can specify the length they need and adjust the pace to fit a particular verse or arrangement.
Why this matters for creatives
For professionals who rely on music production tools, Lyria 3.5 reduces the gap between AI-generated sound and a polished track. The ability to control tempo and duration, combined with more expressive vocals, means the model can serve as a starting point for demos or scratch tracks that require less post-processing. As AI tools continue to evolve, the focus shifts from raw generation to practical creative control - and that's where the latest Lyria model lands. The update is available now in Google Flow Music.
Your membership also unlocks:
