AI Music Extender: make any track longer
Upload a song, pick the second you want it to carry on from, and get a continuation that stays in the same key, tempo and style — two versions per run.
What it does
Continuation, not looping
The usual way to make a track longer is to loop it — copy a section and paste it again. It works for background beds and gives away immediately that nothing new happened. An extender writes new material instead: it listens to what you uploaded, works out the key, tempo, instrumentation and feel, and composes a section that follows on from it.
You choose where the handover happens. Set it to the last second and the track simply keeps going. Set it earlier and everything after that point is rewritten, which is the trick for replacing an ending you never liked.
Because the model is composing rather than copying, each run gives you two different continuations. Neither is the "correct" one — you listen and keep whichever goes where you wanted.
How it works
How to use the AI music extender
Runs take about three to four minutes.
- 01
Upload the track
MP3, WAV, FLAC or M4A up to 20MB. It can be a finished song, a rough demo, or a thirty-second idea.
- 02
Choose where to continue from
Give it a timestamp in seconds. Everything before that point is kept exactly as uploaded; everything after it is newly written.
- 03
Pick from two continuations
Each run returns two takes. Play both, download the one that works, and extend again from the new ending if you want to keep going.
Hear it
Where the original stops and the AI begins
Forty-five seconds straight through: the last fifteen of the original, then thirty the model wrote. The join is at 0:15 — see if you can hear it.
Seamless — original into AI continuation
join at 0:15
Before and after, separately
15s + 30sWhat you get
It has to sound like the same song
Continue from any second
The handover point is yours to set. Extend from the end to add a new section, or from the middle to rewrite an ending — the audio before the split is untouched.
Key, tempo and timbre carried over
The new section is written against what the model hears in your upload, not against a fresh prompt. That is what stops the join sounding like two songs glued together.
Leave the style blank and it follows what it hears. Fill it in and the new section changes direction.
Two takes every run
A continuation is a creative choice, and the first attempt is not always the right one. You get two to compare without paying twice.
Every run returns two different continuations, so you can pick the one that fits instead of paying again to reroll.
Extend repeatedly
Feed the result back in and continue again. A thirty-second sketch can become a full arrangement a section at a time, with you steering at each step.
Who uses it
When a track stops too soon
A demo that never got finished
You have a strong sixteen bars and no idea what comes next. Get two suggestions and take the better one.
Background music for a longer video
Your edit runs three minutes and the track runs ninety seconds. Extend it rather than looping it and hoping nobody notices.
Replacing an ending
Extend from before the outro and write a different one — a proper fade, a hard stop, or another chorus.
Stretching an intro
Need more room before the vocal comes in for a voiceover or a title sequence. Extend from early and keep the new opening.
Getting past a writing block
Two machine-written continuations are often enough to show you what you actually wanted to do next.
Podcast and stream beds
Turn a short cue into something long enough to talk over without an obvious loop point.
FAQ
Music extender questions
Usually not, but it depends on where you cut. Continuing from a natural boundary — the end of a bar, the end of a phrase, a downbeat — gives the cleanest handover. Splitting in the middle of a held note or a vocal word is the case most likely to sound seamed.
Each run adds a section rather than a fixed number of seconds, typically bringing the result to somewhere between ninety seconds and two minutes. If you need more than that, extend the result again — there is no limit on how many times you can continue.
It is a timestamp in seconds into your uploaded file. Audio before it is preserved unchanged; audio after it is discarded and rewritten. Setting it to roughly the length of your file means "just keep going from the end".
Yes. Leave the style field empty and the model follows what it hears in your upload. Fill it in — "add strings", "drop to half tempo", "instrumental outro" — and the new section moves in that direction while staying in the same key and tempo.
Technically yes, and the tool does not check. The rights position does not change, though: extending someone else’s recording creates a derivative of it, so releasing or monetising the result needs their permission. Your own recordings and tracks you generated here are the uncomplicated cases.
MP3, WAV, FLAC, M4A, AAC and OGG, up to 20MB. If your file is larger than that, exporting a lower bitrate MP3 of the relevant part is usually enough.
Yes — tick the instrumental option and the continuation is written without vocals. Leave it unticked on a track that already has singing and the new section will generally keep singing.
Yes — download the MP3 from Suno and upload it here like any other file. Tracks generated on this site can be extended directly from your history without re-uploading.
New accounts come with enough credits for a first extension. Beyond that it is pay-per-run, and every run gives you two continuations for the price of one.
About three to four minutes. You can leave the page while it works; finished runs also show up under History.
Autres outils
Les autres outils de musique IA de Rewave
Give your track a second half
Upload a song and continue it from wherever you like.