Google Lyria 3.5
Google's music model for full-length songs with structure, vocals or instrumentals, guided by a text prompt, optional lyrics and reference images.
Lyria 3.5 is a cloud music-generation model from Google. It composes original music from a text description. It also supports seed control, with output as MP3 and WAV. It runs through Google and Replicate using your own API key, from about $0.08 per generation.
- Released
- Sep 2026
- Pricing
- $0.08 per generation
- Type
- Music generation
- Output formats
- MP3, WAV
- Seed control
- Supported
Samples
Generated with Lyria 3.5 via Replicate: the same briefs given to every music model in the catalog, so the only thing that changes between two models’ clips is the model. The sung brief is only given to models that can sing.
Texture and restraint
Slow ambient piece for a rainy morning: warm analog pads, soft tape hiss, and a simple four-note piano motif repeating. No drums.
Arrangement and dynamic build
Cinematic orchestral build: low strings and a distant horn, tension rising over eight bars into a full brass swell with timpani.
Rhythm, groove, and low end
Downtempo electronic groove at 96 BPM: dusty breakbeat, deep sub bass, a filtered Rhodes chord stab, and light vinyl crackle.
Sung lyrics, diction, and melody
Upbeat indie pop song at 104 BPM with a warm female lead vocal, strummed acoustic guitar, hand claps, and a light kick drum.
[Verse] Window open wide and the kettle on Sunlight on the table where the night has gone [Chorus] Hold on to the morning, hold on to the light Sing it out together, everything feels right
Google and Google DeepMind build the Gemini family of multimodal models, the Imagen and Nano Banana image models, the Lyria music models, and the Veo video models.
deepmind.google ↗Google's most capable Gemini 4 reasoning model for the hardest coding, agentic, mathematical and scientific work, with text, image, audio and video input and a shared 1M-token context.
Google's fast, lower-cost Gemini 4 model for high-volume coding, agentic and multimodal work, with text, image, audio and video input and a shared 1M-token context.
Google's Gemini 4 tier between Hydrogen and Carbon, for complex coding, knowledge work and science at a lower price, with text, image, audio and video input and a shared 1M-token context.
Google's flagship reasoning model for complex software engineering, deep mathematical and scientific problem solving and autonomous multi-step agents, with a shared 1M-token context.
Google's most capable Flash model, with clear gains over 3.7 Flash in software engineering, agentic tasks, and multi-step reasoning, plus vision and a 1M-token context window.
Google's most efficient multimodal Flash model: an updated reasoning stack with stronger coding and computer-use quality, at the same speed and scale as the rest of the Flash line.