Skip to main content
POST
/
producer
/
audios
Audio Generation
curl --request POST \
  --url https://api.mountsea.ai/producer/audios \
  --header 'Content-Type: application/json' \
  --data '
{
  "action": "<string>",
  "model": "<string>",
  "sounds": [
    {}
  ],
  "lyrics": "<string>",
  "title": "<string>",
  "lyricsStrength": 123,
  "weirdness": 123,
  "seed": 123,
  "makeInstrumental": true,
  "audioId": "<string>",
  "coverStrength": 123,
  "coverArtUrl": "<string>",
  "continueAt": 123,
  "extendDuration": 123,
  "replaceStart": 123,
  "replaceEnd": 123
}
'
{
  "taskId": "<string>"
}
The unified audio operation endpoint. Use the action parameter to specify the operation type.

Request Body

action
string
required
The operation type to perform.
ActionDescription
create_musicCreate new music from scratch
extend_musicContinue existing music
cover_musicTransform music style
replace_musicReplace a section of music
swap_music_vocalsReplace vocals
swap_music_soundReplace instrumental
music_variationCreate variations
model
string
required
The model to use for generation.
ModelDescription
FUZZ-2.0 ProLatest flagship model
FUZZ-2.0Standard 2.0 model
FUZZ-2.0 RawRaw output model
FUZZ-1.1 ProProfessional 1.1
FUZZ-1.1Standard 1.1
FUZZ-1.0 ProProfessional 1.0
FUZZ-1.0Standard 1.0
FUZZ-0.8Legacy model
sounds
array
Array of style/sound prompts.
  • create_music: Max 4 items, supports negative strength (-1 to 1) and time range
  • Other actions: Max 1 item, strength 0.2 to 1
Each item contains:
  • text (string): Prompt describing the style
  • strength (number): Prompt strength
  • timeStart (number): Start time in seconds (create only)
  • timeEnd (number): End time in seconds (create only)
lyrics
string
Lyrics content with section tags like [Verse], [Chorus], etc.
title
string
Song title (max 80 characters).
lyricsStrength
number
Lyrics influence strength (0-1). Higher values follow lyrics more closely.
weirdness
number
Creativity/uniqueness level (0-1). Higher values produce more creative results.
seed
number
Random seed for reproducible generation.
makeInstrumental
boolean
default:"false"
Generate instrumental only (ignores lyrics).
audioId
string
Source audio ID. Required for: extend_music, cover_music, replace_music, swap_music_vocals, swap_music_sound, music_variation
coverStrength
number
Cover transformation strength (0-1). Only for: cover_music
coverArtUrl
string
Cover art image URL for the generated song.
continueAt
number
Start position in seconds to continue from. Only for: extend_music
extendDuration
number
Duration in seconds to extend. Only for: extend_music
replaceStart
number
Start time of section to replace. Required for: replace_music
replaceEnd
number
End time of section to replace. Required for: replace_music

Response

taskId
string
The task ID. Use this to query task status via Get Task Status.

Examples

curl -X POST https://api.mountsea.ai/producer/audios \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer your-api-key" \
  -d '{
    "action": "create_music",
    "model": "FUZZ-2.0 Pro",
    "title": "Summer Vibes",
    "sounds": [
      { "text": "upbeat pop with bright synths", "strength": 0.7 }
    ],
    "lyrics": "[Verse]\nSunshine on my face...\n\n[Chorus]\nThis is summer...",
    "lyricsStrength": 0.5,
    "weirdness": 0.3
  }'

Response Example

{
  "taskId": "abc123-def456-ghi789"
}
See the Introduction page for complete examples of each action type with detailed parameters.