What is Suno Replace Section?
Suno Replace Section, model ID suno-ai/infill, rewrites one part of a song without touching the rest. It acts on a clip from a previous Suno generation: you pass that clip_id, mark the section with start_s and end_s, and send the new words in lyrics. The model generates audio for that range so it joins the music on both sides, and returns it as a new clip while the original stays as it was. The model version is set with mv, and optional context_start_s and context_end_s control how much of the surrounding song the model listens to. You can also pass tags for style, a title for the new clip, and make_instrumental to fill the section with music only. Ranges outside the clip are rejected before any generation runs. It is the fix for a song that is almost right: one weak verse, a chorus that needs a different hook, or an intro that drags.
Key features of Suno Replace Section
Five features cover what a replace-section request controls.
Replace any range by seconds
start_s and end_s are required and accept decimals. Everything outside that range stays the same, so you keep the parts of the song that already work.
New lyrics for the section
Pass lyrics with the words for the replaced part. Leave it out, or set make_instrumental to true, to fill the range with music only.
Context window control
context_start_s and context_end_s set how much surrounding audio the model listens to. The defaults are start_s and the end of the clip.
Range checks before generation
A range that falls outside the clip is rejected before the model runs, so a bad timestamp from your UI fails fast instead of producing a broken clip.
Three v6 model versions
Set mv to chirp-hawk, chirp-hawk-wild, or chirp-goose. Add tags to keep the new section in the style you want and title to name the edited clip.
Best for
Fixing one weak verse
Rewrite a single verse without regenerating a song that is otherwise finished.
Swapping a chorus hook
Try several new chorus lines on the same backing and keep the best one.
Localised lyrics
Replace a section with translated lyrics for a regional version of a track.
Better intros and outros
Regenerate the first or last few bars when the opening drags or the ending stops abruptly.
Bar-aligned rewrites
Line start_s and end_s up with bar boundaries for the cleanest joins.
Use cases
A lyric editor can show the song text with timestamps, let the user select a line, type new words, and send the selection as start_s, end_s, and lyrics. A brand jingle tool can keep one approved backing track and swap only the line that names the product for each client. A localisation pipeline can replace the chorus of a song with translated lyrics for each market while the verses stay the same. If you want to cut a section out rather than rewrite it, the Suno Crop API removes a range without generating new audio. New source clips come from Suno music generation.
API examples
All Suno edit endpoints are documented on one page: Suno edit docs.
curl -X POST https://api.unifically.com/v1/tasks \
-H "Content-Type: application/json" \
-H "Authorization: Bearer YOUR_API_KEY" \
-d '{
"model": "suno-ai/infill",
"input": {
"clip_id": "1901ec96-72c6-4c26-8e56-91028ca6070d",
"mv": "chirp-hawk",
"start_s": 15,
"end_s": 35,
"lyrics": "Hello darkness, my old friend"
}
}'
The response returns a task_id. Poll GET /v1/tasks/<task_id> until the edited clip is ready, or include a callback_url outside input to get it by webhook.
FAQs
People also ask
It regenerates a chosen time range inside an existing Suno clip. You set the range with start_s and end_s, pass new lyrics for that part, and the model writes a new section that fits the music around it. The result is a new clip, and the original stays untouched.
Yes. The API model ID is suno-ai/infill. Replace Section is the name used in the playground and docs.
Any seconds inside the clip work, and decimals are accepted. Ranges that fall outside the clip are rejected before generation. Sections that start and end on bar boundaries blend best.
Yes. Leave lyrics empty, or set make_instrumental to true.
They tell the model how much of the song around the section to take into account. By default the context starts at start_s and runs to the end of the clip. Widen it when the new section should follow a melody that starts earlier.
Yes. It is the one Suno edit that generates new audio, so it is billed per request. Crop, fade, speed, and reverse are currently not billed. See the pricing page for the live rate.
