What is Cantai?
Cantai is a machine learning singing instrument trained from real, professional vocalists who gave full consent for their voices to be used. Each singer was contracted, paid for their work and receives royalties for their contributions.
Cantai includes:
Solo voices — Soprano, Alto, Tenor, and Baritone, with stylistic variations (e.g. soft, operatic)
Choir voices — Female Choir and Male Choir
Full integration inside MuseScore Studio
It is the first AI singing instrument fully integrated with music notation software.
How does Cantai work? Is it a machine or samples?
Cantai uses machine learning trained from real human singers, not synthetic voice creation. These singers chose to share their voices for creative use and receive ongoing royalties when Cantai is used.
Cantai is designed to support creativity — not replace it. It is a tool for learning, composing, arranging, rehearsing, and experimenting with voice parts inside MuseScore Studio.
Will more voices be added?
Yes. Cantai will continue to evolve with new voices, improved realism, and more expressive control over time.
Can Cantai be used for more than vocal composition?
Absolutely. Cantai is great for:
Learning and practicing choral parts
Soloing or muting parts during rehearsal
Teaching and ear training
Trying alternate voicings and harmonies
Licensing, Pricing, and Access
What does Cantai cost?
Cantai - MuseSounds Edition is available as a subscription for $14.99/month or as a one time purchase for $299. This gives you access to Cantai voices inside MuseScore Studio.
Does the subscription include web or cloud access outside MuseScore Studio?
No. Cantai - MuseSounds Edition currently works only inside MuseScore Studio 4.6.4 and higher.
Can I use Cantai for published or commercial work?
Yes. You are licensed to use Cantai for personal or professional projects, including sold recordings, provided you own the rights to the music or it is public domain.
Is there a usage limit?
Yes. Cantai includes 180 minutes of audio rendering per week.
After reaching the limit, previously rendered parts will continue to play unless you make changes (dynamics, articulations, notes, etc.).
Usage resets every Monday.
To maximize your Cantai usage, see this article.
Requirements
What do I need to run Cantai?
You will need:
A computer that meets the minimum requirements for MuseScore Studio and MuseHub
The latest version of MuseHub and MuseScore Studio 4.6.4 or higher
An active internet connection to communicate with Cantai servers
Troubleshooting
Cantai downloaded but isn’t showing in MuseScore Studio. What should I do?
Try launching both MuseHub and MuseScore Studio up to three times — in some cases this completes the registration process.
Also make sure:
MuseScore Studio is version 4.6.4 or higher
MuseHub is fully updated
MuseHub remains open while using MuseScore Studio
Data Privacy
Cantai only sends the minimum encrypted notation and data necessary to generate audio.
We do not train new models using your data, and we do not sell or share user data beyond what is required for the system to function. Cantai’s privacy policy is available here: https://www.cantai.app/privacy-policy
Features & Capabilities
What vocal styles does Cantai support?
Classical
Operatic
Choral
Broadway / Light Pop
(Children’s voices are not yet available.)
What languages are currently supported for lyrics?
English
Latin
More languages will be added in future updates.
Does Cantai follow dynamics and expressive markings?
Yes. Cantai responds to dynamics and articulations within MuseScore Studio, and phrasing is determined by each voice model.
Can Cantai be used with other MuseSounds libraries?
Yes. Cantai - MuseSounds Edition works alongside all MuseSounds libraries in the regular MuseScore playback system.
What audio formats can I export?
Cantai follows MuseScore Studio’s standard export formats:
MP3, WAV, OGG, and FLAC
Integration with MuseScore Studio
How do I use Cantai in my score?
Open the Mixer
Select a Cantai voice for the staff
Enter lyrics as usual:
Select a note → Press Ctrl/⌘ + L
Notes will initially play as piano — then switch to vocals once rendered.
Detailed instructions are available here: https://handbook.musescore.org/sound-and-playback/musesounds/online-sounds#how-to-set-up-online-sounds
Do I need to export MIDI or MusicXML first?
No. Cantai works directly inside MuseScore Studio — no external exporting is required.
Recommended Workflow: Switch to Manual Rendering
To control when rendering happens and to save rendering minutes:
Open Preferences
Go to Audio & MIDI
Scroll to Online sounds
Uncheck Automatically process online sounds in the background
The processing indicator at the bottom of MuseScore Studio becomes a button. Click it whenever you want to render.
Limitations & Future Plans
Can Cantai sound like a famous singer? Can I request a custom voice?
Not yet. Cantai does not currently support celebrity voice cloning — but custom voices may be explored in the future.
Support
Where can I report bugs or get help?
Help and online manuals are available here: https://handbook.musescore.org/sound-and-playback/musesounds/online-sounds#how-to-set-up-online-sounds
Contact support here: https://support.musehub.com/hc/en-gb/requests/new