Lyria 3.5 in Public Preview: Next-Gen Music Generation Model Hits the Gemini API

What shipped

Lyria 3.5 — Google's next-generation music generation model (lyria-3.5) — is now in public preview on the Gemini API, announced September 3, 2026.

What changed

Lyria 3.5 generates full-length songs with improved musical coherence and natural vocals, plus fine-grained control over duration and structure. It accepts text and image inputs and outputs high-fidelity 44.1 kHz stereo audio. It's callable through the Gemini API and Google AI Studio, following the standard lyria-3.5 model name — no separate endpoint or provider account.

Why a builder cares

Music generation is now just another Gemini API call: same key, same SDK, same billing as text and vision. That makes Lyria 3.5 an easy add for jingle generators, ambient-score pipelines, and audio-first apps that previously needed a dedicated music provider.