Motifs1.3.0 - App Store

Describe the song you have in mind, in a few words, and motifs makes the whole thing: the music, the lyrics, the voice singing them, and the cover. Your first song is free.

You don't need to sing, play an instrument, or write lyrics. You write what you want to hear - a mood, a story, someone you love, a joke between friends - and press generate. A few minutes later you have a finished song, with its own title and cover. You can listen while it is being made, and if you allow notifications, we tell you the moment it is ready.

WHO IT'S FOR

motifs is for people who want to make a song and share it with someone close to them. A song for a friend's birthday, a song for a couple, a running joke turned into a chorus. It is not a professional production tool and it is not a music catalogue: there is no public feed, and your songs stay private to your account.

PICK A PRODUCER, OR DON'T

Before you generate, you can pick a producer. These are real people who agreed to take part: each wrote the style behind their own pack, their name is used with their permission, and they earn a share of what their pack makes. Every producer has a sound of their own - the instruments, the era, the mood - and it shapes how your song sounds, never what it says.

SHARE A SNIPPET

Choose fifteen to sixty seconds of your song and motifs turns it into a vertical video: the cover, the title, and the lyrics lighting up word by word in time with the singing. Ready to post.

HOW IT WORKS

motifs is an intermediary between you and several AI models, and would rather say plainly which one does what. The music is generated by Treblo. The cover is generated by FLUX.1 [schnell] on fal.ai. Claude orchestrates: it blends the producer's style into your idea, writes the song's title, and writes the description for the cover.

WHAT WE PUT IN PLACE

Before anything is generated, the words you write are checked: if they ask to imitate a named real artist or to redo an existing song, the app asks you to rephrase - to describe a style, a mood, an era instead - and nothing is generated. Genres, eras, instruments and the subject of your song are never affected.

AI-generated content is labelled as such in the app. Producers are real, consenting and paid - and never imitated: their name is never sent to the models, only the style they wrote. Your songs are private. And a generation that fails is always refunded.

No subscription: you buy tokens when you need them.

Version History A new welcome that says what motifs is for: songs to give. Example ideas that type themselves out, a conversation where a song lands on someone's mum, and your first song still free. Producers are now real people, paid a share of their pack. And shared files carry an "AI generated" label in their name. 1.3.0 Aug 24 Songs and covers now carry an "AI generated" label. Lyrics can be copied in one tap. A new About page explains which AI models do what. And if your request names a real artist or song, the app offers to rephrase it. 1.2.0 Aug 21 Partage un extrait de ta chanson en vidéo. Choisis le passage, regarde les paroles s'illuminer mot à mot, et poste-le où tu veux. Et des corrections un peu partout. 1.1.0 Aug 18 1.0 Aug 18