Google Launches Lyria 3.5 Music Model in Flow Music
On July 29, 2026, Google rolled out Lyria 3.5 inside Google Flow Music—its newest music generation model with claimed gains in musicality, lyrics, vocal expression, and direct tempo/duration control for full-song creation.
TLDR
Google on July 29, 2026 launched Lyria 3.5 in Google Flow Music, positioning it as the platform’s next music-generation model. The Google Labs primary highlights four product claims: richer musicality, higher-quality lyrics with better prompt adherence, more expressive vocals, and easier tempo and duration control. Availability is Flow Music (web product surface), not a standalone third-party model API at announcement.
What shipped
| Item | Detail (Google primary) |
|---|---|
| Model | Lyria 3.5 |
| Surface | Google Flow Music |
| Musicality | Richer, more complex melodic structures |
| Lyrics | Higher quality; improved prompt adherence and structural awareness |
| Vocals | More realistic / emotionally nuanced; improved pronunciation |
| Controls | Tempo and duration of outputs |
| Watermarking | SynthID lineage continues for generative audio (platform practice; confirm per-output in product) |
Google frames Flow Music as the consumer create surface: generate songs you can control and iterate, not only one-shot clips. Secondary coverage (Android Central, AI Music Preneur) notes Lyria 3.5 replacing Lyria 3 Pro as Flow Music’s generation backend on the same day.
Competitive context
AI music remains a multi-vendor race (Suno, Udio, Meta Muse-adjacent audio, other lab audio stacks). Google did not publish head-to-head benches against Suno/Udio on the Labs post; early secondary takes still treat paid Suno tiers as quality peers. The differentiated Google pitch is end-to-end Flow (create → edit controls → Google product distribution) plus brand trust/watermarking, not open weights.
Why this story matters
Music generation is a consumer multimodal lane that labs treat as strategically important for engagement and creative tooling. A named Lyria version bump with tempo/duration control signals Google is iterating productized music models, not only research demos. Watch whether Lyria 3.5 reaches developer APIs, quality vs Suno on independent boards, and how Flow Music monetizes generation volume.