Lyria 3.5
Google DeepMind Lyria 3.5 — music from a description.
Try it
You can fill this in now — signing in happens when you run it, and your inputs come with you.
How to use Lyria 3.5
- Fill in Song description in the form on this page.
- Adjust any option in the form — the defaults are the platform's recommended values.
- Run it; signing in starts the charge, and the result (Text to music) appears above the form, ready to download.
At a glance
- Developer
- Can do
- Text to music
- Credits per run
- 8 credits
- Typical time
- ~60s
- Released
- 2026-07-29
- Added
- 2026-09-26
What you can adjust
Everything in the form above and what each can be set to. The defaults are the platform's recommended values.
- Song descriptionText you write
- Negative promptText you write
What Lyria 3.5 costs
Lyria 3.5 costs about 8 credits a run, usually back in around 60 seconds. Credits are shared across every workspace, new accounts start with free ones, and a failed run is refunded.
Lyria 3.5 compared with other Music models
Same workspace, different trade-offs — credits, speed and what each one can be asked for.
| AI models | Can do | Credits per run | Typical time |
|---|---|---|---|
| Lyria 3.5 | Text to music | 8 | ~60s |
| MiniMax Music 3.0 | Text to music | 9 | ~60s |
| Sonilo v1.1 | Text to music | 18 | ~60s |
| Stable Audio 3.0 | Text to music | 3 | ~60s |
| MiniMax Music 2.6 | Text to music | 12 | ~60s |
Questions
What can Lyria 3.5 do?
Text to music. Fill in the form on this page to try it.
How many credits does Lyria 3.5 cost?
About 8 credits per run, charged at the platform's live price when you run it. One credit balance covers every workspace.
How long does it take?
Around 60 seconds; heavier settings or a queue take longer, and the page shows progress.
Do I need an account, and who owns the output?
Browsing and filling in the form need no account; running it does (new accounts start with free credits). What you generate is yours, personally or commercially, subject to the terms.
Other models like this
Where this model is used: Music·All models