Audio Cleaner (AudioSeparation) API - Integration Guide

Overview

AudioSeparation is an API for source separation and vocal isolation. It can intelligently split tracks such as lead vocals, backing vocals, and accompaniment from songs, making it suitable for karaoke production, remixing, and practice accompaniment use cases.

Key Features

• One-click vocal removal: quickly separate vocals and accompaniment for common karaoke needs.
• Multi-track outputs: export target stems like vocals and instrumental for post-production.
• High-quality fidelity: preserves musical details while reducing distortion and artifacts.
• Broad format support: accepts many mainstream audio/video formats from common sources.
• Efficient processing: async task workflow suitable for batch jobs and automated pipelines.

Authentication

All API requests require api-key using an API key. Include your API key in the request header:

API Host: https://audiocleaner.ai/

api-key: YOUR_API_KEY

Pricing Rules

This audio separation service is billed by task audio duration, rounded up to whole minutes.

Each task reports `duration` (seconds), and the backend automatically calculates billable minutes.

1. Create Audio Separation Task

Endpoint

POST /audio/api/v1/api-keys/separation/task/create

Request Headers

Request Parameters (Body · JSON)

Available model_type Options

Parameter Example

{
  "model_type": 350,
  "inputurl": "https://resource.audiocleaner.ai/acweb/audio/audio_splitter.mp4"
}

Response Format (Unified Response)

{
  "code": 100000,
  "data": {
    "job_id": "33b6251d7ba94edea8361c318736e746"
  },
  "message": "Request Success"
}

Detailed Field Descriptions

• code: Numeric response status code. Common values: 100000 (success), 100100 (insufficient credits), 101002 (invalid API key), 101003 (rate limit), 101004 (concurrency limit), 101005 (API key disabled), 101006 (API key not enabled).
• data: Result payload, varies by scenario. Example fields include job_id (unique task ID) and output (separation result URL).
• message: Human-readable message describing the current status.

Main Response Codes & Examples - 100000: Request Success

{
  "code": 100000,
  "data": {
    "job_id": "33b6251d7ba94edea8361c318736e746"
  },
  "message": "Request Success"
}

100100: Insufficient Credits

{
  "code": 100100,
  "data": {},
  "message": "Insufficient credit balance"
}

101002: Invalid API Key

{
  "code": 101002,
  "data": {},
  "message": "Invalid API key"
}

101003: API Rate Limit Exceeded

{
  "code": 101003,
  "data": {},
  "message": "API key rate limit exceeded"
}

101004: Concurrent Request Limit Exceeded

{
  "code": 101004,
  "data": {},
  "message": "API key concurrent request limit exceeded"
}

101005: API Key Disabled

{
  "code": 101005,
  "data": {},
  "message": "Your account is suspended. Status cannot be modified. Please contact the administrator"
}

101006: API Key Not Enabled

{
  "code": 101006,
  "data": {},
  "message": "API key not enabled"
}

Notes and Recommendations

• All endpoints return `code`, `data`, and `message` for easier frontend handling and display.
• When credits are insufficient or the API key is abnormal, `data` is usually an empty object `{}` with a clear `message`.
• Common error scenarios are covered. Contact platform support if you still have questions.

Example Request

curl --location --request POST 'https://audiocleaner.ai/audio/api/v1/api-keys/separation/task/create' \
  --header 'Content-Type: application/json' \
  --header 'api-key: YOUR_API_KEY' \
  --data-raw '{
    "model_type": 350,
    "inputurl": "https://resource.audiocleaner.ai/acweb/audio/audio_splitter.mp4"
  }'

2. Query Task Status (New Model)

Endpoint

POST /audio/api/v1/api-keys/separation/task/get

Request Headers

Request Parameters

Example Request

curl --location --request POST \
 'https://audiocleaner.ai/audio/api/v1/api-keys/separation/task/get' \
  --header 'Content-Type: application/json' \
  --header 'api-key: YOUR_API_KEY' \
  --data-raw '{
    "job_id": "YOUR_JOB_ID"
  }'

Response Field Descriptions

output Field vs. Model Mapping (Based on AudioSeparation Component)

Response Code Reference

Response Example

{
  "code": 100000,
  "data": {
    "status": "waiting/success/failed",
    "file_type": "audio",
    "video_url": "https://resource.audiocleaner.ai/paid/source-file/202604/14/1776146445_before.wav",
    "output": {
      "vocals": "https://resource.audiocleaner.ai/paid/output/result/202604/14/1776146482_before_vocals.wav",
      "other": "https://resource.audiocleaner.ai/paid/output/result/202604/14/1776146483_before_other.wav"
    },
    "zip_url": "https://resource.audiocleaner.ai/paid/output/result/202604/14/1776146484_before.zip"
  },
  "message": "Request Success"
}

Use Cases

• Karaoke accompaniment production: quickly remove lead vocals and keep background music.
• Remixing and derivative creation: extract vocals or accompaniment for rearrangement.
• Music practice and teaching: get clean accompaniment tracks for instrumental or vocal training.
• Batch separation processing: run one async task per audio file for automation-friendly workflows.

FAQ