The simplest way to use AudioLasso: create a key, submit a job, wait for it to finish, and download the result.

AudioLasso lets you pull one sound out of an audio or video file using a plain-English prompt.

The API follows a queue model: submit work, store the returned request_id, poll status, then fetch the result. Streaming and webhooks are optional alternatives for clients that need them.

If you only learn one thing, learn this workflow:

Create an API key.
Submit a separation job.
Wait for the job to finish.
Download the result files.

1. Create an API key

Open API keys, create a key, and copy it once.

Create an API key

Store it in your shell:

export AUDIOLASSO_API_KEY="al_..."

Every API request sends it as a bearer token:

Authorization: Bearer $AUDIOLASSO_API_KEY

2. Submit a job

If your file is already hosted at a public HTTPS URL, you can submit it directly with input.audio_url.

curl -X POST https://audiolasso.dev/v1/queue/audio/separate \
  -H "Authorization: Bearer $AUDIOLASSO_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "input": {
      "audio_url": "https://example.com/audio.wav",
      "prompt": "isolate the lead vocal"
    }
  }'

AudioLasso replies immediately with a queued job. Store the request_id if another process needs to check the same job later.

{
  "request_id": "req_01jabc",
  "status": "IN_QUEUE",
  "model": "audio/separate",
  "status_url": "https://audiolasso.dev/v1/queue/requests/req_01jabc/status",
  "stream_url": "https://audiolasso.dev/v1/queue/requests/req_01jabc/status/stream",
  "result_url": "https://audiolasso.dev/v1/queue/requests/req_01jabc/result"
}

3. Wait for it to finish

Polling is the canonical wait path:

curl https://audiolasso.dev/v1/queue/requests/req_01jabc/status \
  -H "Authorization: Bearer $AUDIOLASSO_API_KEY"

You only need to care about four states:

Status	Meaning
`IN_QUEUE`	The request is waiting to start
`IN_PROGRESS`	The model job is running
`COMPLETED`	Outputs are ready
`FAILED`	Processing failed

4. Download the result

When the job reaches COMPLETED, fetch the signed output URLs:

curl https://audiolasso.dev/v1/queue/requests/req_01jabc/result \
  -H "Authorization: Bearer $AUDIOLASSO_API_KEY"

{
  "request_id": "req_01jabc",
  "model": "audio/separate",
  "status": "COMPLETED",
  "data": {
    "target": {
      "url": "https://...",
      "content_type": "audio/wav",
      "file_name": "target.wav"
    },
    "residual": {
      "url": "https://...",
      "content_type": "audio/wav",
      "file_name": "residual.wav"
    },
    "sample_rate": 48000
  },
  "usage": {
    "feature": "audio_seconds",
    "quantity": 42,
    "billing_status": "deducted"
  }
}

JavaScript example

const apiKey = process.env.AUDIOLASSO_API_KEY;

const submit = await fetch("https://audiolasso.dev/v1/queue/audio/separate", {
  method: "POST",
  headers: {
    Authorization: `Bearer ${apiKey}`,
    "Content-Type": "application/json",
  },
  body: JSON.stringify({
    input: {
      audio_url: "https://example.com/audio.wav",
      prompt: "isolate the lead vocal",
    },
  }),
});

if (!submit.ok) throw new Error(await submit.text());
const job = await submit.json();

while (true) {
  const statusRes = await fetch(job.status_url, {
    headers: { Authorization: `Bearer ${apiKey}` },
  });
  const status = await statusRes.json();

  if (status.status === "FAILED") throw new Error("Audio separation failed");
  if (status.status === "COMPLETED") break;

  await new Promise((resolve) => setTimeout(resolve, 3000));
}

const resultRes = await fetch(job.result_url, {
  headers: { Authorization: `Bearer ${apiKey}` },
});
const result = await resultRes.json();

console.log(result.data.target.url);
console.log(result.data.residual.url);

Python example

import os
import time
import requests

api_key = os.environ["AUDIOLASSO_API_KEY"]
headers = {"Authorization": f"Bearer {api_key}"}

submit = requests.post(
    "https://audiolasso.dev/v1/queue/audio/separate",
    headers={**headers, "Content-Type": "application/json"},
    json={
        "input": {
            "audio_url": "https://example.com/audio.wav",
            "prompt": "isolate the lead vocal",
        }
    },
)
submit.raise_for_status()
job = submit.json()

while True:
    status = requests.get(job["status_url"], headers=headers)
    status.raise_for_status()
    payload = status.json()

    if payload["status"] == "FAILED":
        raise RuntimeError("Audio separation failed")
    if payload["status"] == "COMPLETED":
        break

    time.sleep(3)

result = requests.get(job["result_url"], headers=headers)
result.raise_for_status()
data = result.json()

print(data["data"]["target"]["url"])
print(data["data"]["residual"]["url"])

Local files

If your file is local or private, upload it first with POST /v1/files, then submit the returned input.file_id.

See Upload a file.

Quickstart