Lyria 3.5 AI Music Generator

AudioLyria 3.5Duration
Create now

Google DeepMind's Lyria 3.5. Full songs with vocals and lyrics.

Every AI music model

One workspace
Lyria 3.5Google DeepMind
Seed Audio 1.0ByteDance
ElevenLabsElevenLabs

Made with Lyria 3.5

Lo-fi study beat

Instrumental, Rhodes and vinyl

Cinematic build

Orchestral, piano to full brass

Indie pop, sung

Vocals, written lyrics

Dark trap beat

Instrumental, 808s in D minor

French chanson

Vocals, prompted in French

Chiptune loop

Instrumental, retro 8-bit

Generate now

How to generate a song with Lyria 3.5

  1. 01

    Open Morphic

    Sign up and start creating on a free-flowing, infinite visual canvas.

  2. 02

    Describe the track

    Genre, instruments, BPM, key, and mood. Add section tags or timestamps to set the shape.

  3. 03

    Generate and refine

    A full song comes back at 44.1 kHz stereo. Adjust the prompt and swap models on the Canvas.

Generate now

What Lyria 3.5 generates

Scores that land on the cut

Write a timeline into the prompt and the arrangement follows it. Name what plays from 0:00 to 0:10, then from 0:10 to 0:30, and the swell arrives on the edit instead of near it.

A darkened scoring stage with a conductor lit by a single warm key light

Songs from lyrics you wrote

Paste your verses and choruses under their section tags and hear them performed as a finished arrangement, with vocals that carry the phrasing rather than reciting the words.

A handwritten lyric notebook open beside an acoustic guitar in warm lamplight

Instrumental beds for anything

Game menus, explainer backing, podcast themes, and ad beds. Ask for instrumental only and the track leaves room for the dialogue that sits on top of it.

A dim studio desk with a mixing surface and a glowing waveform on a dark monitor

Music composed from your images

Attach up to ten stills and the model scores what it sees, reading mood, palette, and setting. Feed it frames from your timeline and the music matches the grade.

A wall of framed cinematic film stills arranged in a grid above a studio desk

Anatomy of a Lyria 3.5 prompt

Name four things and the model has what it needs: the genre it is playing, the instruments carrying it, the tempo and key it sits in, and the shape it follows.
GenreInstrumentsTempo and keyStructure
GenreA lo-fi hip hop beatInstrumentswith dusty vinyl crackle, mellow Rhodes chords, an upright bass line,Tempo and keya slow boom-bap pattern at 85 BPM,Structureinstrumental only, 30 seconds
GenreAn upbeat feel-good pop songInstrumentswith bright acoustic guitar, claps, and warm vocal harmonies,Tempo and keyin G major at 120 BPM,Structure[Verse] [Chorus] [Verse] [Chorus]
GenreAn epic cinematic orchestral pieceInstrumentsopening on solo piano, building through sweeping strings to massive brass,Tempo and keyslow and deliberate, around 70 BPM,Structure[0:00 - 0:20] Intro [0:20 - 1:00] Build
Generate now

Related workflows

Put yourself in a movie

Drop in a photo of yourself. Get a 12-second cinematic movie with animation and a matched soundtrack.

Try this workflow
Messages to cinematic scene

Messages to cinematic scene

Dramatize any text conversation into a cinematic multi-shot video.

Try this workflow

Renovation timelapse

Photo of any real-estate becomes a cinematic renovation timelapse video with construction SFX and music.

Try this workflow

Time-freeze effect

Cinematic time-freeze short film for any scene

Try this workflow
Music video

Music video

Pick a visual style, drop in your lyrics. Each verse gets its own scene visual, laid out as a full screenplay.

Try this workflow
Cinematic storyboard

Cinematic storyboard

Share a scene description with character references. Get a full storyboard with shot angles and mood.

Try this workflow

All on Morphic

Your complete audio stack

Lyria 3.5

The full model page: song structure, timestamped control, image input, and the limits worth knowing before you start.

Lyria 3.5 prompt guide

The six things worth naming in every prompt, how to write a timeline, and the mistakes that get a prompt blocked.

Voice and dialogue

Narration, character voices, and scripted conversation. Cast from a description, a character image, or a reference clip.

Workflows

Turn any repeatable audio job into a one-click rerun. Save it as a Workflow, then share or rerun it anytime.

Simple pricing

Get started for free today, with the option to upgrade or cancel anytime.

Basic

$9/ month
billed as $0 per year

1100 monthly credits

1 user only

All models

Workflows

Standard

$24/ month
billed as $0 per year

3625 monthly credits

1 user only

All models

Workflows

Pro

$45/ month
billed as $0 per year

6350 shared monthly credits

1 user

+ up to 4 more at extra cost

All models

Workflows

Pro Max

$170/ month
billed as $0 per year

24650 shared monthly credits

1 user

+ up to 9 more at extra cost

All models

Workflows

Enterprise

For higher limits

Custom

pricing and billing terms

High-volume credits
Custom seat limits
All models
Workflows
Pricing Gradient

Free

For playing around

$0

forever free

Up to 20 credits
1 user only
Limited models
Workflows
Compare plan details

FAQs

What is the Lyria 3.5 AI music generator?
Lyria 3.5 is Google DeepMind's music generation model. Describe the track you want and it returns a full-length song, around two minutes, with an arrangement it works out first: an intro, verses, a chorus, a bridge. Vocals and lyrics come with it, and the audio arrives at 44.1 kHz stereo, the same sample rate as a CD master.
How is this different from other AI music generators?
Two things. The model reasons through the musical structure before it renders any audio, so a song holds together across two minutes instead of looping one idea. And you can write a timeline into the prompt, naming what happens between 0:00 and 0:10 and again from 0:10 to 0:30, so instruments enter where you put them rather than where the model guesses.
Can I use my own lyrics?
Yes. Paste them into the prompt and mark the sections with [Verse], [Chorus], and [Bridge] tags. Keep the lyrics as their own block, separate from your musical direction, so the model sings your words instead of reading them as a description. Leave lyrics out and it writes them for you, then hands them back as text alongside the song.
Can it generate instrumental music with no vocals?
Yes. Ask for it directly, as in "instrumental only, no vocals". It is the usual route for game audio, background beds, podcast themes, and anything that has to sit underneath dialogue.
What languages can it sing in?
The prompt language sets the lyric language. Write your brief in French and the song comes back sung in French, with the vocal style and pronunciation adapted to it. There is no separate setting to change, so write the whole prompt in the language you want to hear.
Can I use the Lyria 3.5 AI music generator online?
Yes. Everything runs in the browser on Morphic, with no download or install. Write the prompt, generate, and the track lands on your Canvas from any laptop, including machines with no GPU.
Is the Lyria 3.5 AI music generator free?
Morphic offers a free tier so you can start generating music right away, no payment required. For higher volume or commercial use, paid plans are available.

You might also like