Suno API clients: Python and Node examples

Before you start

• Keep the key server-side. Never ship it in front-end code, an app bundle or a public repo.
• Generation is asynchronous. A successful submit returns a task; the audio appears later.
• Budget per generation, not per request. One call returns two tracks.
• Read-only calls are free - status, downloads, lyrics and MIDI are not billed.

Python

import time, requests

BASE = "https://www.suno-api.io"
H = {"Authorization": "Bearer YOUR_API_KEY"}

# 1) submit
song = requests.post(f"{BASE}/api/music/create", headers=H, timeout=300, json={
    "description": "warm lo-fi piano loop", "model": "suno-v6"
}).json()["data"]["clips"][0]["song_id"]

# 2) poll until complete
for _ in range(60):
    status = requests.post(f"{BASE}/api/music/query", headers=H, timeout=60,
                           json={"song_id": song}).json()
    if status.get("data", {}).get("status") in ("complete", "completed"):
        break
    time.sleep(4)

# 3) fetch a media url (free)
url = requests.post(f"{BASE}/api/music/download-url", headers=H, timeout=60,
                    json={"song_id": song, "kind": "song"}).json()
print(url)

Node

const BASE = "https://www.suno-api.io";
const H = { Authorization: "Bearer YOUR_API_KEY", "Content-Type": "application/json" };

const submit = await fetch(`${BASE}/api/music/create`, {
  method: "POST", headers: H, body: JSON.stringify({ description: "warm lo-fi piano loop", model: "suno-v6" })
}).then((r) => r.json());

const songId = submit.data.clips[0].song_id;
let status = "";
while (status !== "complete" && status !== "completed") {
  const q = await fetch(`${BASE}/api/music/query`, {
    method: "POST", headers: H, body: JSON.stringify({ song_id: songId })
  }).then((r) => r.json());
  status = q.data?.status ?? "";
  if (status !== "complete" && status !== "completed") await new Promise((r) => setTimeout(r, 4000));
}

The four mistakes worth avoiding

• A 30-second client timeout. Generation can take minutes; treat a timeout as "still running" and query the task.
• Resubmitting after a timeout. That is how duplicate charges happen - the first order was already placed.
• No backoff on 429. Add exponential backoff instead of retrying immediately.
• Storing the audio URL. Download links are temporary; store the song id and fetch a fresh URL when you need the file.