Start free

AI Music Generator

Describe the song, the mood or the scene it scores, and pick the model to make it: Eleven Music, Lyria, MiniMax Music and more, with vocals or without. Every track on this page came out of Leap.

Model

Prices are for one track at the length each model starts on. Start free, no card needed; your idea waits for you in the studio after you sign in, and you add lyrics there.

Afrobeats: out of the city, by the seaEleven Music v2.5
0 / 30 s

Prompt

Afrobeats song with warm male vocals, log drums, bright highlife guitar, shakers, about leaving the city for a weekend by the sea

Make one like it

Made on Leap, one prompt each

Real runs, as they came back, each with its prompt, its lyrics when it had them, and the model that made it: songs with vocals, film score, dance music and folk.

  • Country pop: Old Red won't startMiniMax Music 2.6
    0 / 76 s

    Prompt

    Country pop, bright female vocal, pedal steel, acoustic guitar, upbeat, small-town humor

    Read the lyrics

    [Verse] My truck won't start on Sunday It only runs when I'm late The preacher's waving from the porch The whole town's at the gate [Chorus] Old red, don't fail me now I swear I'll wash you twice Old red, don't fail me now Just get me there tonight

    Make one like it
  • Synthwave for an empty coastal highwayLyria 3.5
    0 / 60 s

    Prompt

    Instrumental synthwave for a night drive on an empty coastal highway: pulsing arpeggiated bass, gated snare, warm analog lead, neon and wet asphalt

    Make one like it
  • 90s R&B: say it in the group chatMiniMax Music 2.6
    0 / 62 s

    Prompt

    90s R&B slow jam, smooth male vocal with falsetto, finger snaps, lush pads, round bass

    Read the lyrics

    [Verse] You left me on read at midnight Typing, then nothing at all Three dots like a heartbeat Then the screen goes dark [Chorus] Say it in the group chat, baby Let the whole world see Say it in the group chat, baby That you're coming back to me

    Make one like it
  • Bossa nova, a lazy afternoonLyria 3
    0 / 31 s

    Prompt

    Instrumental bossa nova: nylon-string guitar, brushed drums, warm upright bass, soft alto flute melody, a lazy sunny afternoon, no vocals

    Make one like it
  • A Balkan brass band in 7/8Stable Audio 3 Medium
    0 / 30 s

    Prompt

    Balkan brass band, fast 7/8 rhythm, tuba bassline, trumpets trading melodies, a wedding that won't end

    Make one like it
  • Flamenco guitar and palmasACE-Step
    0 / 45 s

    Prompt

    flamenco, nylon guitar rasgueado, fast palmas hand claps, cajón, passionate, Spanish guitar, instrumental

    Make one like it
  • A tense string quartetLyria 2
    0 / 33 s

    Prompt

    Tense minimalist string quartet: a repeating pizzicato ostinato, cello swells underneath, building suspense, cinematic

    Make one like it
  • Deep house at sunriseSonilo Music 1.1
    0 / 30 s

    Prompt

    Deep house: warm pads, rolling bassline, crisp hi-hats, a soulful hum, 122 bpm, sunrise on a rooftop after the party

    Make one like it
  • Drum and bass, 174 bpmStable Audio 2.5
    0 / 30 s

    Prompt

    Drum and bass, 174 bpm, rolling Reese bass, chopped breakbeat, dark and fast, warehouse at 4 a.m.

    Make one like it

Pick the model for the track

Some models sing the lyrics you write, some write their own, and some make instrumentals only. One account runs them all, so you can try an idea on two and keep the better track.

  • Eleven Music v2.5$0.66 for a 30 s track

    Studio-quality songs with vocals or instrumentals, from a prompt.

    From a prompt, up to 240 s. By ElevenLabs.

    Try Eleven Music v2.5
  • MiniMax Music 3$0.132 for a 60 s track

    Songs up to five minutes long from a style prompt and your lyrics. Without lyrics it makes an instrumental. A song can end before the length you ask for, which is what the run is billed.

    Sings your lyrics, up to 300 s. By MiniMax.

    Try MiniMax Music 3
  • Lyria 3.5$0.11 for a track

    Google's newest music model, from a prompt or an image.

    From a prompt. By Google.

    Try Lyria 3.5
  • Lyria 3$0.044 for a track

    Google's Lyria 3: 30-second tracks at a low price.

    From a prompt. By Google.

    Try Lyria 3
  • ACE-Step$0.0132 for a 60 s track

    Open-weight songs from genre tags and lyrics, priced by the second.

    Sings your lyrics, up to 240 s. By ACE Studio.

    Try ACE-Step
  • Stable Audio 3 Medium$0.04136 for a 30 s track

    Stereo music up to six minutes long, trained on licensed data.

    From a prompt, up to 380 s. By Stability AI.

    Try Stable Audio 3 Medium
See every voice and sound model

What it costs

Start free, then pay per track

A new account comes with $2 of credit, and no card is needed. The table shows each model's price for one track at the length its form starts on, and about how many tracks $2 makes. Longer tracks cost more, and a run that fails costs nothing.

ModelPriceWhat $2 makes
Eleven Music v2.5$0.66for a 30 s trackAbout 3 tracks
MiniMax Music 3$0.132for a 60 s trackAbout 15 tracks
Lyria 3.5$0.11for a trackAbout 18 tracks
Lyria 3$0.044for a trackAbout 45 tracks
ACE-Step$0.0132for a 60 s trackAbout 151 tracks
Stable Audio 3 Medium$0.04136for a 30 s trackAbout 48 tracks

Each price is one track at the length the model's form starts on, from the live catalog; the studio shows the exact price before every run. How pricing works.

Put it to work

Add a voice and video

A track usually scores something. Make the voice-over and the pictures in the same studio.

AI voice generator
Text to speech with Eleven v3, MiniMax Speech and Gemini, for the words over your track.
AI video generator
Make the clip your track scores, with Veo 3.1, Kling 3.0 and Seedance.

Questions

Questions people ask

Is it free?
You can start free: a new account comes with $2 of credit, and no card is needed. That is about 151 tracks with ACE-Step ($0.0132 for a 60 s track), or 3 with Eleven Music v2.5 ($0.66 for a 30 s track). After that you add credit when you want more and pay per track. There is no subscription, and the welcome credit is one per person.
Can it sing my own lyrics?
Yes, with MiniMax Music 3 and ACE-Step: paste your lyrics into the model's form in the studio, with [Verse] and [Chorus] marking the parts as the songs on this page do, and describe the style in the prompt. Eleven Music v2.5 writes its own words from your prompt, as in the Afrobeats track on this page.
How long can a track be?
Up to 380 seconds with Stable Audio 3 Medium. Most models let you pick the length, and a longer track costs more; Lyria 3 makes 30-second tracks. The price on the Generate button changes with the length you pick.
Can I make an instrumental?
Yes. Say so in the prompt, or switch on the instrumental setting where a model has one, as Eleven Music v2.5 does. MiniMax Music 3 makes an instrumental when you give it no lyrics.
Can I use the music commercially?
Leap's Terms say that, as between you and Leap, you own the results the studio makes for you, to the extent the law allows, and they set no personal-use-only limit. In some countries AI-made music may not be protected by copyright. Write your own lyrics and describe a style rather than naming an artist: the Acceptable Use Policy forbids prompting with lyrics you have no right to use and impersonating anyone. If the training data matters to you, Stability AI says Stable Audio 3 Medium was trained on licensed data.
What happens if a run fails?
It costs nothing. Only a run that succeeds is charged, and you see the price on the Generate button before you press it.

Your first song, from one line.

Start free, no card needed. Every run is priced before you press Generate.

Make a song