Full songs with vocals
Turn a short brief into a complete track with a sung lead, harmonies, and a mastered stereo mix you can publish as it is.
output: 2:30 songAI Music Generator turns a sentence, a set of lyrics, or a reference into a finished track with vocals, harmonies, and a mastered stereo mix. Two takes per prompt, downloadable as MP3, WAV, or separated stems.
Describe the song in plain language, or switch to Custom mode and bring your own lyrics. The studio returns two versions of the same idea with vocals, instrumentation, and a mastered mix, and you keep the one that fits.
Write the song you hear in your head the way you would explain it to a friend: the mood, the instruments, the story, and the line you want in the chorus. The studio turns that brief into a full arrangement instead of a short loop.
Switch to Custom mode when you already know what you want and steer the take until it fits.
Every take comes back as a mixed, mastered stereo file that is yours to keep, with the parts still separable if you want to work on them.
Generate only what you have the rights to use. AI Music Generator is a studio tool: you are responsible for the lyrics you submit, for any voice you ask it to work with, and for how you publish what you make.
The same studio handles a full song with a sung lead, a wordless instrumental bed, your own lyrics, and the separated parts of something you already generated.
Turn a short brief into a complete track with a sung lead, harmonies, and a mastered stereo mix you can publish as it is.
output: 2:30 songGenerate background music for a video, a podcast, or a stream, with nothing in the way of your own voice.
output: instrumentalWrite your own words in Custom mode, or take the draft the studio suggests and rewrite it line by line.
output: lyricsPull a finished song apart into vocals, drums, bass, and other parts, then rebuild the mix your own way.
output: 4 stemsAdd another verse past the first ending, sing an old take in a new style, or remix it without starting over.
output: new takeA prompt studio is only useful if the take that comes back is finished enough to publish. These are the parts that make that possible.
Describe the mood, the story, or the exact chorus you want, and the studio answers with a complete arrangement rather than a loop you have to assemble yourself.
A lead vocal with backing harmonies sits on top of the mix, or you can ask for a clean instrumental when you would rather sing it yourself.
Paste your own words in Custom mode, keep the line breaks, and steer the delivery with style tags so the phrasing follows your writing.
Genre, mood, tempo, and instrument tags shape every take, so two prompts about the same subject never come back sounding the same.
Each prompt returns two different versions of the same idea, so you can keep the one that fits and drop the one that does not.
Continue a song past its first ending, or rework an existing take into a cover in another style without losing the melody you liked.
Pull vocals, drums, bass, and other parts apart and download them as separate files for your editor or your DAW.
Download a finished track in the format your workflow wants, without a watermark and without a queue that resets your credits.
Paid plans include the rights you need to publish your songs, monetise them, and keep what you earn from them.
Choose the newest model for the best vocals and the longest songs, or an earlier version whose sound you simply prefer.
Match the feel of a reference track you own, or keep one voice consistent across a whole set of songs in the same project.
Your prompts, lyrics, and songs stay in your account. Nothing is published, and nothing is shared, until you decide it should be.
No plugins, no install, and no blank page to stare at. Write the brief, listen to two takes, and keep the one that works.
Type the idea, paste lyrics, or point at a reference. Add style tags if you already know where the track should sit.
The studio returns two different versions in a couple of minutes. Play them back, keep the one you like, extend it if it ends too soon.
Grab the MP3, the WAV, or the stems, then publish the track under the rights your plan includes.
Pick a style to get close on the first prompt, then bend it with mood, tempo, and instrument tags until the take sounds like the song you had in mind.
Newer versions sound better and write longer songs, older versions are cheaper and faster. Nothing stops you from drafting on one and finishing on another.
| Version | Song length | Details |
|---|---|---|
| v4.5 | Up to 4:30 per track | The newest generation: the most natural vocals, the cleanest mix, and the widest range of styles. Longest songs and the highest credit cost. |
| v4 | Up to 3:00 per track | Strong vocals and a faster queue at a lower credit cost, which is why most everyday tracks are still made on this version. |
| v3.5 | Up to 2:00 per track | The cheapest version and the quickest way to test an idea. Best for drafts you plan to regenerate on a newer model. |
Most sessions start with a deadline, a demo that needs a voice, or an idea that has been sitting in a notes app for a year.
Intros, outros, and background beds you can publish without hunting down a licence or a stock library subscription.
Hear your lyrics sung before you book studio time, then keep the stems and rebuild the arrangement with real players.
Short tracks for ads, social posts, and product videos, with commercial rights included on every paid plan.
Turn a half-formed idea into a real arrangement you can play along with, or practise producing on separated stems.
Monthly and annual plans include song credits, the newest model, stem downloads, and commercial rights. Cancel whenever you like, and keep every song you already made.
The details people ask about most often before they generate their first song.
A complete song from a written prompt: vocals, harmonies, instrumentation, and a mastered stereo mix, usually in a couple of minutes. You can also ask for an instrumental, a cover of an earlier take, an extension past the first ending, or the separated stems of something you already generated.
Paid plans include commercial rights, so you can publish the tracks, monetise them, and keep what you earn. On the free allowance the tracks are for personal, non-commercial listening and testing only.
A credit is the unit the studio charges per generation, and a single prompt spends credits once and returns two takes. Longer songs on newer model versions cost more credits than short drafts on older ones. Your plan refills credits every month, and annual plans refill them for the whole year.
Yes. Switch to Custom mode, paste your lyrics with the line breaks you want, and add style tags to steer the delivery. You can also start from the draft the studio suggests and edit it before you generate again.
It depends on the model version: up to about four and a half minutes on the newest one, three minutes on the version before it, and two minutes on the cheapest one. Any take can be extended past its ending if you want a longer arrangement.
Yes. A finished song can be split into its vocal, drum, bass, and other parts, and each part can be downloaded separately. Stems are useful when you want to re-balance the mix, replace a part, or practise producing on a real arrangement.
On a paid plan, yes: the commercial rights in your plan cover publishing, monetising, and licensing the songs you generate. You are still responsible for the lyrics you submit and for not imitating a real artist or a protected work.
No. The studio runs in the browser on a desktop or a phone, and your library of songs is tied to your account rather than to one device. Sign-in is through Google, so there is no separate password to manage.
Open the billing section of your account and change or cancel the subscription there. Cancelling stops the next payment, keeps the credits you have already been given for the current period, and leaves every song you already downloaded in your hands.
Sign in with Google, describe the song, and keep the finished mix, with the lyrics and the stems included.