lyria-3.5 — Music Generation AI Model
Lyria writes music from a description: Lyria 3.5 composes a full song with verses, chorus and vocals, Lyria 3 Clip a 30-second loop. An image can set the mood. Returns an MP3 and the lyrics. Runs on Google's side against your API key.
What it does
Lyria turns a written description - genre, instruments, tempo, mood, and the words to sing if you have them - into a 44.1 kHz stereo MP3. Lyria 3.5 writes a full song whose length follows the prompt; Lyria 3 Clip always returns a 30-second clip.
An image on the optional port is read as a mood board, and songs with vocals also come back with their lyrics on a text port. Nothing runs locally; it needs a Google AI key.
When to use it
- A background track for a short-form video, matched to its mood
- A full song with vocals from a theme and a few lines of lyrics
- Loopable 30-second cues for previews and drafts on Lyria 3 Clip
- Music that follows the look of an image wired in from earlier
Details
- CategoryMusic Generation
- Year2026
- LicenseGoogle Cloud Terms
Compliance & Provenance
- ProviderGoogle · GPAI
- EU AI Act RiskLimited Risk
- Art. 50 TransparencyRequired — AI-generated outputs are marked
Inputs & Outputs
- TextInputstring
Description of the music: genre, instruments, mood, tempo, and lyrics to sing
- ImageInputimageoptional
Optional image whose mood the music should follow
- SoundOutputsound
Generated music (MP3)
- LyricsOutputstring
Lyrics of the generated song (empty for instrumentals)
Parameters
- modelstringdefault Lyria 3.5options: Lyria 3.5, Lyria 3 Clip
Model variant
'Lyria 3.5' (default) — full-length song, about a couple of minutes, length steerable in the prompt. 'Lyria 3 Clip' — always a 30-second clip; half the price, for loops and previews.
- promptstringdefault
Style / extra instruction
Optional instruction prepended to the text input, e.g. 'instrumental only, 90 BPM lo-fi'. Leave blank to use the text input alone.
Limitations
- No seed, negative prompt or exact duration field - say what to avoid and how long it should run in the prompt itself.
- Lyria 3 Clip always returns 30 seconds regardless of what the prompt asks for.
- Runs on your own Google AI API key. Without one configured in the workspace the node cannot run.
Tags
Alternatives in Audio Models
- Text to Music (ACE-Step 1.5)
ACE-Step 1.5: generate music from a style prompt, with optional lyrics, key, tempo and time signature. 48 kHz stereo, 5-600s.
- Text to Sound Effect (MOSS-SoundEffect v2)
MOSS-SoundEffect v2.0: generate a sound effect from an English description. 48 kHz mono, 0.5-30s.
- Text to Speech (Gemini)
Google Gemini TTS: turn text into a spoken voiceover (Korean/English). Needs API key.
- Text to Speech (Qwen3 Voice Design)
Qwen3-TTS VoiceDesign: on-device text-to-speech with a voice you describe in words. No API key.
- Text to Speech (Qwen3)
Qwen3-TTS: on-device text-to-speech voiceover (Korean/English). No API key.