Background Removal API

Remove the background
of any image*
with 1 API call

$ curl -H 'X-API-Key: YOUR_API_KEY'                      \
       -F '[email protected]/path/to/file.jpg'                \
       -f https://api.remove.bg/v1.0/removebg -o no-bg.png

$ curl -H 'X-API-Key: YOUR_API_KEY' \
  -F '[email protected]/path/to/file.jpg' \
  -f https://api.remove.bg/v1.0/removebg \
  -o no-bg.png

Get API Key

* Any image that has a foreground (tested with people, products, animals, cars, etc.)

Easy to integrate

Our API is a simple HTTP interface with various options:

Get started

  1. Get your API Key.
    Your first 50 API calls per month are on us (see Pricing).
  2. Use the following code samples to get started quickly
  3. Review the reference docs to adjust any parameters

Sample Code

Remove background from
Environment
Code 
$ curl -H 'X-API-Key: INSERT_YOUR_API_KEY_HERE'           \
       -F '[email protected]/path/to/file.jpg'                 \
       -F 'size=auto'                                     \
       -f https://api.remove.bg/v1.0/removebg -o no-bg.png

// Requires "request" to be installed (see https://www.npmjs.com/package/request)
var request = require('request');
var fs = require('fs');

request.post({
  url: 'https://api.remove.bg/v1.0/removebg',
  formData: {
    image_file: fs.createReadStream('/path/to/file.jpg'),
    size: 'auto',
  },
  headers: {
    'X-Api-Key': 'INSERT_YOUR_API_KEY_HERE'
  },
  encoding: null
}, function(error, response, body) {
  if(error) return console.error('Request failed:', error);
  if(response.statusCode != 200) return console.error('Error:', response.statusCode, body.toString('utf8'));
  fs.writeFileSync("no-bg.png", body);
});

# Requires "requests" to be installed (see python-requests.org)
import requests

response = requests.post(
    'https://api.remove.bg/v1.0/removebg',
    files={'image_file': open('/path/to/file.jpg', 'rb')},
    data={'size': 'auto'},
    headers={'X-Api-Key': 'INSERT_YOUR_API_KEY_HERE'},
)
if response.status_code == requests.codes.ok:
    with open('no-bg.png', 'wb') as out:
        out.write(response.content)
else:
    print("Error:", response.status_code, response.text)

# Install "remove_bg" first (https://github.com/remove-bg/ruby)
require "remove_bg"

RemoveBg.from_file("/path/to/file.jpg",
  api_key: "INSERT_YOUR_API_KEY_HERE"
).save("no-bg.png")

// Requires "guzzle" to be installed (see guzzlephp.org)

$client = new GuzzleHttp\Client();
$res = $client->post('https://api.remove.bg/v1.0/removebg', [
    'multipart' => [
        [
            'name'     => 'image_file',
            'contents' => fopen('/path/to/file.jpg', 'r')
        ],
        [
            'name'     => 'size',
            'contents' => 'auto'
        ]
    ],
    'headers' => [
        'X-Api-Key' => 'INSERT_YOUR_API_KEY_HERE'
    ]
]);

$fp = fopen("no-bg.png", "wb");
fwrite($fp, $res->getBody());
fclose($fp);

// Requires "Apache HttpComponents" to be installed (see hc.apache.org)

Response response = Request.Post("https://api.remove.bg/v1.0/removebg")
    .addHeader("X-Api-Key", "INSERT_YOUR_API_KEY_HERE")
    .body(
        MultipartEntityBuilder.create()
        .addBinaryBody("image_file", new File("/path/to/file.jpg"))
        .addTextBody("size", "auto")
        .build()
    ).execute();
response.saveContent(new File("no-bg.png"));

// Requires AFNetworking to be installed (see https://github.com/AFNetworking/AFNetworking)

NSURL *fileUrl = [NSBundle.mainBundle URLForResource:@"file" withExtension:@"jpg"];
NSData *data = [NSData dataWithContentsOfURL:fileUrl];
if (!data) {
    return;
}

AFHTTPSessionManager *manager =
[[AFHTTPSessionManager alloc] initWithSessionConfiguration:
 NSURLSessionConfiguration.defaultSessionConfiguration];

manager.responseSerializer = [AFImageResponseSerializer serializer];
[manager.requestSerializer setValue:@"INSERT_YOUR_API_KEY_HERE"
                 forHTTPHeaderField:@"X-Api-Key"];

NSURLSessionDataTask *dataTask = [manager
                                  POST:@"https://api.remove.bg/v1.0/removebg"
                                  parameters:nil
                                  constructingBodyWithBlock:^(id  _Nonnull formData) {
                                      [formData appendPartWithFileData:data
                                                                  name:@"image_file"
                                                              fileName:@"file.jpg"
                                                              mimeType:@"image/jpeg"];
                                  }
                                  progress:nil
                                  success:^(NSURLSessionDataTask * _Nonnull task, id  _Nullable responseObject) {
                                      if ([responseObject isKindOfClass:UIImage.class] == false) {
                                          return;
                                      }

                                      self.imageView.image = responseObject;
                                  } failure:^(NSURLSessionDataTask * _Nullable task, NSError * _Nonnull error) {
                                      // Handle error here
                                  }];

[dataTask resume];

// Requires Alamofire to be installed (see https://github.com/Alamofire/Alamofire)

guard let fileUrl = Bundle.main.url(forResource: "file", withExtension: "jpg"),
    let data = try? Data(contentsOf: fileUrl) else { return }

Alamofire
    .upload(
        multipartFormData: { builder in
            builder.append(
                data,
                withName: "image_file",
                fileName: "file.jpg",
                mimeType: "image/jpeg"
            )
        },
        to: URL(string: "https://api.remove.bg/v1.0/removebg")!,
        headers: [
            "X-Api-Key": "INSERT_YOUR_API_KEY_HERE"
        ]
    ) { result in
        switch result {
        case .success(let upload, _, _):
            upload.responseJSON { imageResponse in
                guard let imageData = imageResponse.data,
                    let image = UIImage(data: imageData) else { return }

                self.imageView.image = image
            }
        case .failure:
            return
        }
    }
using (var client = new HttpClient())
using (var formData = new MultipartFormDataContent())
{
    formData.Headers.Add("X-Api-Key", "INSERT_YOUR_API_KEY_HERE");
    formData.Add(new ByteArrayContent(File.ReadAllBytes("/path/to/file.jpg")), "image_file", "file.jpg");
    formData.Add(new StringContent("auto"), "size");
    var response = client.PostAsync("https://api.remove.bg/v1.0/removebg", formData).Result;

    if(response.IsSuccessStatusCode) {
        FileStream fileStream = new FileStream("no-bg.png", FileMode.Create, FileAccess.Write, FileShare.None);
        response.Content.CopyToAsync(fileStream).ContinueWith((copyTask) =>{ fileStream.Close(); });
    } else {
        Console.WriteLine("Error: " + response.Content.ReadAsStringAsync().Result);
    }
}

$ curl -H 'X-API-Key: INSERT_YOUR_API_KEY_HERE'           \
       -F 'image_url=https://www.remove.bg/example.jpg'   \
       -F 'size=auto'                                     \
       -f https://api.remove.bg/v1.0/removebg -o no-bg.png

// Requires "request" to be installed (see https://www.npmjs.com/package/request)
var request = require('request');
var fs = require('fs');

request.post({
  url: 'https://api.remove.bg/v1.0/removebg',
  formData: {
    image_url: 'https://www.remove.bg/example.jpg',
    size: 'auto',
  },
  headers: {
    'X-Api-Key': 'INSERT_YOUR_API_KEY_HERE'
  },
  encoding: null
}, function(error, response, body) {
  if(error) return console.error('Request failed:', error);
  if(response.statusCode != 200) return console.error('Error:', response.statusCode, body.toString('utf8'));
  fs.writeFileSync("no-bg.png", body);
});

# Requires "requests" to be installed (see python-requests.org)
import requests

response = requests.post(
    'https://api.remove.bg/v1.0/removebg',
    data={
        'image_url': 'https://www.remove.bg/example.jpg',
        'size': 'auto'
    },
    headers={'X-Api-Key': 'INSERT_YOUR_API_KEY_HERE'},
)
if response.status_code == requests.codes.ok:
    with open('no-bg.png', 'wb') as out:
        out.write(response.content)
else:
    print("Error:", response.status_code, response.text)

# Install "remove_bg" first (https://github.com/remove-bg/ruby)
require "remove_bg"

RemoveBg.from_url("https://www.remove.bg/example.jpg",
  api_key: "INSERT_YOUR_API_KEY_HERE"
).save("no-bg.png")

// Requires "guzzle" to be installed (see guzzlephp.org)

$client = new GuzzleHttp\Client();
$res = $client->post('https://api.remove.bg/v1.0/removebg', [
    'multipart' => [
        [
            'name'     => 'image_url',
            'contents' => 'https://www.remove.bg/example.jpg'
        ],
        [
            'name'     => 'size',
            'contents' => 'auto'
        ]
    ],
    'headers' => [
        'X-Api-Key' => 'INSERT_YOUR_API_KEY_HERE'
    ]
]);

$fp = fopen("no-bg.png", "wb");
fwrite($fp, $res->getBody());
fclose($fp);

// Requires "Apache HttpComponents" to be installed (see hc.apache.org)

Response response = Request.Post("https://api.remove.bg/v1.0/removebg")
    .addHeader("X-Api-Key", "INSERT_YOUR_API_KEY_HERE")
    .body(
        MultipartEntityBuilder.create()
        .addTextBody("image_url", "https://www.remove.bg/example.jpg")
        .addTextBody("size", "auto")
        .build()
    ).execute();
response.saveContent(new File("no-bg.png"));

// Requires AFNetworking to be installed (see https://github.com/AFNetworking/AFNetworking)

AFHTTPSessionManager *manager =
[[AFHTTPSessionManager alloc] initWithSessionConfiguration:
 NSURLSessionConfiguration.defaultSessionConfiguration];

manager.responseSerializer = [AFImageResponseSerializer serializer];
[manager.requestSerializer setValue:@"INSERT_YOUR_API_KEY_HERE"
                 forHTTPHeaderField:@"X-Api-Key"];

NSURLSessionDataTask *dataTask = [manager
                                  POST:@"https://api.remove.bg/v1.0/removebg"
                                  parameters:@{@"image_url": @"https://www.remove.bg/example.jpg"}
                                  progress:nil
                                  success:^(NSURLSessionDataTask * _Nonnull task, id  _Nullable responseObject) {
                                      if ([responseObject isKindOfClass:UIImage.class] == false) {
                                          return;
                                      }

                                      self.imageView.image = responseObject;
                                  } failure:^(NSURLSessionDataTask * _Nullable task, NSError * _Nonnull error) {
                                      NSLog(@"ERROR");
                                  }];
[dataTask resume];

// Requires Alamofire to be installed (see https://github.com/Alamofire/Alamofire)

Alamofire
    .request(
        URL(string: "https://api.remove.bg/v1.0/removebg")!,
        method: .post,
        parameters: ["image_url": "https://www.remove.bg/example.jpg"],
        encoding: URLEncoding(),
        headers: [
            "X-Api-Key": "INSERT_YOUR_API_KEY_HERE"
        ]
    )
    .responseData { imageResponse in
        guard let imageData = imageResponse.data,
            let image = UIImage(data: imageData) else { return }

        self.imageView.image = image
    }
using (var client = new HttpClient())
using (var formData = new MultipartFormDataContent())
{
    formData.Headers.Add("X-Api-Key", "INSERT_YOUR_API_KEY_HERE");
    formData.Add(new StringContent("https://www.remove.bg/example.jpg"), "image_url");
    formData.Add(new StringContent("auto"), "size");
    var response = client.PostAsync("https://api.remove.bg/v1.0/removebg", formData).Result;

    if(response.IsSuccessStatusCode) {
        FileStream fileStream = new FileStream("no-bg.png", FileMode.Create, FileAccess.Write, FileShare.None);
        response.Content.CopyToAsync(fileStream).ContinueWith((copyTask) =>{ fileStream.Close(); });
    } else {
        Console.WriteLine("Error: " + response.Content.ReadAsStringAsync().Result);
    }
}

Libraries and tools

Output Formats

You can request one of three formats via the format parameter:

Format Resolution Pros and cons Example
PNG Up to 10 Megapixels
e.g. 4000x2500 		+ Simple integration
+ Supports transparency
- Large file size 		Download
7 MB
JPG Up to 25 Megapixels
e.g. 6250x4000 		+ Simple Integration
+ Small file size
- No transparency supported 		Download
1 MB
ZIP Up to 25 Megapixels
e.g. 6250x4000 		+ Small file size
+ Supports transparency
- Integration requires compositing 		Download
3 MB

Please note that PNG images above 10 megapixels are not supported. If you require transparency for images of that size, use the ZIP format (see below). If you don't need transparency (e.g. white background), we recommend JPG.

How to use the ZIP format

The ZIP format has the best runtime performance for transparent images. In comparison to PNG, the resulting file is up to 80% smaller (faster to download) and up to 40% faster to generate. For performance optimization we recommend using the ZIP format whenever possible. Above 10 megapixels, usage of the ZIP format is required for transparent results.

The ZIP file always contains the following files:

color.jpg A non-transparent RGB image in JPG format containing the colors for each pixel. (Note: This image differs from the input image due to edge color corrections.) – Show example
alpha.png A non-transparent grayscale image in PNG format containing the alpha matte. White pixels are foreground regions, black is background. – Show example

To compose the final image file:

  1. Extract the files from the ZIP
  2. Apply the alpha matte (alpha.png) to the color image (color.jpg).
  3. Save the result in a format of your choice (e.g. PNG)

Sample code for linux (requires zip and imagemagick): 

$ unzip no-bg.zip                      && \
  convert color.jpg alpha.png             \
    -compose CopyOpacity                  \
    -composite no-bg.png

A zip2png command-line tool and more code samples can be found here.

API Reference

Rate Limit

You can process up to 500 images per minute through the API, depending on the input image resolution in megapixels.

Examples:

Input image Megapixels Effective Rate Limit
625 x 400 1 MP 500 images per minute
1200 x 800 1 MP 500 images per minute
1600 x 1200 2 MP 500 / 2 = 250 images per minute
2500 x 1600 4 MP 500 / 4 = 125 images per minute
4000 x 2500 10 MP 500 / 10 = 50 images per minute
6250 x 4000 25 MP 500 / 25 = 20 images per minute

Exceedance of rate limits leads to a HTTP status 429 response (no credits charged). Clients can use the following response headers to gracefully handle rate limits:

Response header Sample value Description
X-RateLimit-Limit 500 Total rate limit in megapixel images
X-RateLimit-Remaining 499 Remaining rate limit for this minute
X-RateLimit-Reset 1600962581 Unix timestamp when rate limit will reset
Retry-After 59 Seconds until rate limit will reset (only present if rate limit exceeded)

Higher Rate Limits are available upon request.

Exponential backoff

Exponential backoff is an error handling strategy in which a client periodically retries a failed request. The delay increases between requests and often includes jitter (randomized delay) to avoid collisions when using concurrent clients. Clients should use exponential backoff whenever they receive HTTP status codes 5XX or 429.

The following pseudo code shows one way to implement exponential backoff: 

retry = true
retries = 0

WHILE (retry == TRUE) AND (retries < MAX_RETRIES)
  wait_for_seconds (2^retries + random_number)

  result = get_result

  IF result == SUCCESS
    retry = FALSE
  ELSE IF result == ERROR
    retry = TRUE
  ELSE IF result == THROTTLED
    retry = TRUE
  END
    
  retries = retries +1 
END

The maximum number of retries and the delay between retries depends on the use case of your application and network conditions. Mobile clients may need to retry more often and need longer delays compared to desktop clients.

API Changelog

Most recent API updates:

  • 2020-05-06: Added semitransparency parameter.
  • 2019-09-27: Introduce ZIP format and support for images up to 25 megapixels.
  • 2019-09-25: Increased maximum file size to 12 MB and rate limit to 500 images per minute. Rate limit is now resolution-dependent.
  • 2019-09-16: Added enterprise credit balance field to account endpoint.
  • 2019-08-01: Added add_shadow option for car images.
  • 2019-06-26: Added crop_margin, scale and position parameters.
  • 2019-06-19: Added support for animals and other foregrounds (response header X-Type: animal and X-Type: other)
  • 2019-06-11: Credit balances can now have fractional digits, this affects the X-Credits-Charged value
  • 2019-06-03: Added parameters bg_image_url, bg_image_file (add a background image), crop (crop off empty regions) and roi (specify a region of interest).
  • 2019-05-13: Added car support (type=car parameter and X-Type: car response header)
  • 2019-05-02: Renamed size "regular" to "small". Clients should use the new value, but the old one will continue to work (deprecated)
  • 2019-05-01: Added endpoint GET /account for credit balance lookups
  • 2019-04-15: Added parameter format to set the result image format
  • 2019-04-15: Added parameter bg_color to add a background color to the result image

Thanks for your feedback!

Want to help us improve? (optional)

Choose if you want us to use this image for quality improvements.
Choose if you do NOT want us to use this image for quality improvements.

Your image will be used for future improvements of remove.bg.

Your image will NOT be used for future improvements of remove.bg.