AI Music Generator — Turn a Spark into Sound
AI music agents for songs, lyrics, MIDI and live coding
Describe what you want in plain words. Each agent works on the real thing (the song, the lyrics, the notes or the code), and you can still edit the lyrics, notes and code yourself.
AI Music Agent
Describe a song, or start from a hummed melody, a photo or your own recording. The agent drafts two directions and starts on the first, then remixes, extends, adds vocals or a backing track, and masters any track on a visual canvas. Suggested edits run in one click.
AI Lyrics Agent
Chat with an AI songwriter that writes and rewrites your lyrics in an editor beside the chat, a line or a section at a time, then take them to the Music Agent to make a song.
AI MIDI Agent
Ask for a melody or an edit in plain words. The agent writes the notes into the piano roll, and each request can be undone in one click.
Strudel AI Agent
A Chrome side panel for live-coding music on Strudel. It reads your pattern, explains it and rewrites it when you describe the sound you want.
opens the Chrome Web Store in a new tab
One Lacuna account and one credit balance work across every agent.
Your AI Music Team, on One Canvas
Connect every step — writing, arranging, generating, and stem-splitting — on one visual canvas
- One idea
- Two directions
- Four tracks
- Every edit, its own branch
Edit, Remix and Extend Every Track — Without Starting Over
Remix, extend, rewrite a section, split stems, or master — directly from the track node
Change the lyrics of one section with AI
Region Edit is an AI music editor for one section of a song. Select the section on the waveform and write the new lyrics; Lacuna regenerates only that part, and the rest of the song stays as it was. Use it as a song lyrics changer for a few lines, a chorus or a verse, without generating the song again.
- The original lines are pre-filled when lyrics are synced
- A built-in lyrics assistant helps draft the new lines
- Saved as a new version; the original stays on the canvas
Remix a whole song in a new style with AI
Remix is an AI song remixer that reimagines the whole track. Describe the sound you want, or start from style directions suggested for the song, then keep the lyrics or rewrite them. Your original audio serves as the reference, so the AI remix starts from your song, not from scratch.
- The style prompt starts from the song's own style
- Up to three versions per run on higher plans
- Each remix is saved next to the original
Create album art without leaving the track
Generate new album art for the song you're working on. The prompt starts from the track's style, and an AI prompt helper can write one for you. Apply the image you like and it replaces the track's artwork, while the audio stays exactly as it was. For reference images, other aspect ratios or upscaling, use the Album Cover Generator.
- Square artwork, applied to the track in one click
- Example covers you can reuse as a starting prompt
- Results also appear in your Album Cover Generator history
Extend a song with AI from the point you choose
Extend works as an AI music extender: it writes a new section for your song instead of looping what's already there. Write the lyrics for the new part yourself, or let AI continue them from the lines that come before. Use it as an AI song extender for an extra chorus, a bridge or an outro.
- Add a section at the end, or rework the ending
- The new section's length follows its lyrics
- The extended version is saved next to the original
Split a song into vocals, drums, bass and more
Stems uses AI to split the track into two stems (vocals and instrumental), four (vocals, drums, bass and other) or six, adding guitar and piano. Solo, mute, pan and rebalance each part in the mixer, then download the stems or render a version without one part, such as the vocals. To split a file from your computer, use the Stem Splitter or the Vocal Remover.
- Choose between the Lacuna Pro and Lacuna Basic models
- Download each stem as its own MP3
- Removing a part can reuse stems you've already split
Master a track and compare every version
Mastering polishes loudness, dynamics and tone by matching your mix to reference presets such as Balanced & Clear or Loud & Punchy. Each preset you pick becomes its own mastered version, and you can switch between the original and every master on one timeline. To match a reference song of your own, open AI Mastering.
- Readings for LUFS, true peak (dBTP) and loudness range (LU)
- Limiter on by default, or off to keep headroom
- Output as WAV (16- or 24-bit) or 320 kbps MP3
Change the singing voice on any track
Voice cover keeps a song's lyrics, melody, timing and backing track, and changes who sings it. Pick an official voice, or train a private one from 2–30 minutes of your own clean singing with AI Voice Cloning. A ready voice can also cover a song from your computer in the AI Cover Generator, or sing brand-new songs in AI Song Generator with My Voice.
- Lyrics, melody and the backing track stay as they were
- Official voices, or a private voice trained on your singing
- Saved next to the original as a new track
Tune the vocals of any track, from subtle to robotic
Autotune separates the vocals, corrects their pitch to the song's key and mixes them back over the untouched instrumental. Keep it natural with a light touch, or push it to the hard, robotic effect. For a recording from your computer, use Autotune Online.
- Detects the key from the full mix, or set it yourself
- The default keeps your vibrato; stronger presets tame or remove it
- The tuned track is saved next to the original
How to make music with AI
Make a song with AI in four steps, from a first idea to a finished track.
Describe your idea
Tell the Music Agent what you want: a genre, a mood, a story, or lyrics you've already written. Ask for an instrumental if you don't need vocals, or start from a hummed melody, a photo or your own recording. Have finished lyrics? Paste them into Lyrics to Song, or start from a one-line brief with Text to Song.
Get two directions
The agent drafts two directions for your idea, keeping what you asked for and varying what you left open, then starts generating the first. Adjust either one, or ask for something different.
Listen and compare
Play the takes on the canvas as they finish; most models return two per run. For more options, generate the second direction or switch models.
Refine, then release
Edit a section, remix, extend, tune the vocals, change the voice, split stems or master the track, then turn the finished song into a lyric video or a visualizer.
Start from a melody, a photo or your own recording
Words are only one way in. Hum a tune, upload a picture or a MIDI file, or bring a vocal or a beat, and the Music Agent builds the song around it.
- Text to SongDescribe a story, a mood or a moment in a line or two, and get lyrics and a finished song.
- Lyrics to SongPaste lyrics you've written and hear them sung with a melody and a full arrangement.
- Hum to SongHum or sing a tune, check the notes, then grow it into a full song or an instrumental.
- Image to MusicUpload a photo, illustration or screenshot and get an original song or instrumental that fits its mood.
- MIDI to SongAttach a MIDI melody and Lacuna Cadence arranges and sings a full song around it, up to five minutes long.
- Your own recordingBring a vocal and add a backing track, bring a beat and add vocals, or upload a finished song to remix, extend or master it.
Separate, clean up, mix and master the audio you already have
Upload a song or a recording from your computer. These tools run on the same models as the canvas, and you don't need to generate anything first.
Separate vocals and instruments
Split a song into stems, remove the vocals for karaoke, keep the backing vocals when you remove the lead singer, or pull out a single instrument.
Clean up and restore
Reduce reverb and crowd noise, restore detail in compressed music, or remove the music from a video and keep the dialogue and sound effects.
Mix, tune and master
Turn separate stems, or a vocal and a beat, into a stereo mix, tune a vocal, and master the result to release loudness.
Rework a song
Blend two songs into a mashup, change the key or tempo, find the BPM and key, or loop a section to practice.
Words in time, visuals on the beat.
Sync lyrics to any song word by word, turn them into a lyric video, or give the track an audio-reactive visualizer for any screen.
LRC Generator
Turn any song into word-level timed lyrics
Lyric Video Maker
Turn a track and its lyrics into a shareable video
Music Visualizer
Turn a song into an audio-reactive video in 16:9, 9:16 or 1:1.
The Online MIDI Editor That Does It All
A full online MIDI editor — precise control in your browser
Audio, sheet music or a prompt in, editable MIDI out — then back to sheet music, MP3 or WAV.
- MP3 to MIDI
- MIDI to Sheet Music
- Sheet Music to Audio
- MIDI to MP3
- MIDI to WAV
- Sheet Music Scanner
- Sheet Music Simplifier
- Sight Reading Generator
- ABC to MIDI
Explore AI-Generated Music
Every track starts with a simple description — click any piece to hear AI's imagination
Explore Our AI Music Generator & Creative Tools
AI music generation, lyrics writing, and creative tools for every musician.
Create
- AI Song Generator
- AI Lyrics Generator
- Text to Song
- Lyrics to Song
- Hum to Song
- Image to Music
- MIDI to Song
- Sheet Music to Song
- AI Song Extender
- Song Generator with My Voice
- AI Voice Cloning
- AI Cover Generator
- Phonk Maker
- AI Beat Maker
- AI Instrumental Generator
- Story to Song
- Notes to Song
- Rap Lyrics Generator
- Chord Progression Generator
- AI Radio Station
Separate
Clean up
MIDI & Notation
Frequently Asked Questions
Everything you need to know about AI music generation
How does AI music generation work?
Simply describe your musical vision — genre, mood, tempo, instruments — or provide your lyrics. Lacuna's AI analyzes your input and generates a complete, original song with vocals, melody, harmony, and instrumentation in a few minutes.
Can I generate songs with vocals?
Yes! Lacuna generates complete songs with AI vocals. You can specify vocal style, gender, and tone to match your creative vision. The result sounds natural and professional.
What music genres does Lacuna support?
Lacuna supports virtually every genre — pop, rock, R&B, hip-hop, country, classical, jazz, electronic, Latin, folk, and many more. You can also blend genres for unique hybrid sounds.
Who owns the music I generate?
You do! Paid users receive full commercial rights to all generated music and lyrics. Use your creations on YouTube, Spotify, podcasts, games, or any commercial project.
Is it really free to start?
Yes. Every free account gets monthly credits to generate and listen to full songs — no credit card required. When you're ready for more, our flexible plans offer additional credits, downloads and commercial use.
Do I need an account to generate music?
Yes. Sign in to generate and save tracks. Creating an account is free, includes complimentary monthly credits, and does not require a credit card to start.