How to Integrate a Background Removal API (Step by Step)

This is the hands-on companion to our complete API guide. If you want to know exactly how to integrate a background removal API — from zero to a working, production-ready call — this walkthrough takes you through every step using the AutoBackgrounding API (from 5¢ per image).

Prerequisites

• A server-side environment (Node.js, Python, PHP, or anything that can make HTTPS requests).
• An AutoBackgrounding account and API key (covered in Step 1).
• A publicly accessible vehicle image URL, or a local file to upload.

Step 1: Get Your API Key

Sign up and open your dashboard to generate an API key. Keys are secret credentials — store them in an environment variable, never in source control or client-side code. Grab your key from the API pricing & docs page.

"tok-comment"># Store your key as an environment variable
AUTOBG_API_KEY=sk_live_your_key_here

Step 2: Authenticate

Every request includes an Authorization header with your key as a bearer token. That's the entire auth model — no OAuth dance, no signing.

Authorization: Bearer sk_live_your_key_here
Content-Type: application/json

Step 3: Make Your First Call

Send a single vehicle image to POST /v1/process. Here it is in three languages — pick yours:

curl https:"tok-comment">//api.autobackgrounding.com/v1/process \
  -H "Authorization: Bearer $AUTOBG_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "image_url": "https:">//example.com/lot-photo.jpg",
    "background": "studio_white"
  }'

Step 4: Handle the Response

The API responds with JSON. A successful result looks like this:

{
  "id": "job_8f2a1c",
  "status": "completed",
  "output_url": "https:">//cdn.autobackgrounding.com/out/8f2a1c.png",
  "credits_used": 1
}

Read status first. If it's completed, use output_url. If it's failed, inspect the error and note that failed jobs consume zero credits.

Step 5: Download & Store the Output

Don't hotlink the returned URL in production — download the processed image and store it in your own bucket (S3, Vercel Blob, GCS, etc.), then reference your copy.

import { writeFile } from "node:fs/promises";

const img = await fetch(data.output_url);
const buffer = Buffer.from(await img.arrayBuffer());
await writeFile("./processed/8f2a1c.png", buffer);
"tok-comment">// Or upload the buffer to S3 / Vercel Blob here

Step 6: Error Handling & Retries

Production integrations must handle transient failures. Retry 5xx and 429 (rate limited) responses with exponential backoff; treat other 4xx codes as permanent. Use your own reference ID so retries stay idempotent.

async function processWithRetry(payload, attempts = 4) {
  for (let i = 0; i < attempts; i++) {
    const res = await fetch("https:">//api.autobackgrounding.com/v1/process", {
      method: "POST",
      headers: {
        Authorization: `Bearer ${process.env.AUTOBG_API_KEY}`,
        "Content-Type": "application/json",
      },
      body: JSON.stringify(payload),
    });

    if (res.ok) return res.json();

    "tok-comment">// Retry only on rate limit or server errors
    if (res.status === 429 || res.status >= 500) {
      const wait = 2 ** i * 500; "tok-comment">// 500ms, 1s, 2s, 4s
      await new Promise((r) => setTimeout(r, wait));
      continue;
    }

    "tok-comment">// Permanent error
    throw new Error(`API error ${res.status}: ${await res.text()}`);
  }
  throw new Error("Exhausted retries");
}

Next Steps

You now have a working, resilient single-image integration. From here:

• Processing a whole inventory? Move to the bulk / batch endpoint.
• Want push notifications instead of polling? Set up webhooks for async processing.
• Building in Node or Python specifically? Follow the Node.js or Python deep-dive.

Get your API key and the full endpoint reference on the API pricing & docs page — pricing starts at 5¢ per image.