Skip to content
Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time
  Python SDK

Python CI imagekitio codecov License: MIT Twitter Follow

ImageKit Python SDK allows you to use image resizing, optimization, file uploading and other ImageKit APIs from applications written in the Python language.

Supported Python Versions: >=3.6

Table of contents -


Go to your terminal and type the following command

pip install imagekitio


from imagekitio import ImageKit
imagekit = ImageKit(
    private_key='your private_key',
    public_key='your public_key',
    url_endpoint = 'your url_endpoint'


You can use this Python SDK for 3 different kinds of methods - URL generation, file upload, and file management. The usage of the SDK has been explained below.

URL generation

1. Using Image path and image hostname or endpoint

This method allows you to create a URL using the path where the image exists and the URL endpoint(url_endpoint) you want to use to access the image. You can refer to the documentation here to read more about URL endpoints in ImageKit and the section about image origins to understand about paths with different kinds of origins.

imagekit_url = imagekit.url({
            "path": "/default-image.jpg",
            "url_endpoint": "",
            "transformation": [{"height": "300", "width": "400"}],

The result in a URL like,w-400/default-image.jpg

2.Using full image URL This method allows you to add transformation parameters to an absolute URL using src parameter. This method should be used if you have the complete image URL stored in your database.

image_url = imagekit.url({
    "src": "",
    "transformation" : [{
        "height": "300",
        "width": "400"

The results in a URL like

The .url() method accepts the following parameters.

Option Description
url_endpoint Optional. The base URL to be appended before the path of the image. If not specified, the URL Endpoint specified at the time of SDK initialization is used. For example,
path Conditional. This is the path at which the image exists. For example, /path/to/image.jpg. Either the path or src parameter needs to be specified for URL generation.
src Conditional. This is the complete URL of an image already mapped to ImageKit. For example, Either the path or src parameter needs to be specified for URL generation.
transformation Optional. An array of objects specifying the transformation to be applied in the URL. The transformation name and the value should be specified as a key-value pair in the object. Different steps of a chained transformation can be specified as different objects of the array. The complete list of supported transformations in the SDK and some examples of using them are given later. If you use a transformation name that is not specified in the SDK, it gets applied as it is in the URL.
transformation_position Optional. The default value is path that places the transformation string as a path parameter in the URL. It can also be specified as query, which adds the transformation string as the query parameter tr in the URL. If you use the src parameter to create the URL, then the transformation string is always added as a query parameter.
query_parameters Optional. These are the other query parameters that you want to add to the final URL. These can be any query parameters and not necessarily related to ImageKit. Especially useful if you want to add some versioning parameter to your URLs.
signed Optional. Boolean. Default is false. If set to true, the SDK generates a signed image URL adding the image signature to the image URL. This can only be used if you are creating the URL with the url_endpoint and path parameters and not with the src parameter.
expire_seconds Optional. Integer. Meant to be used along with the signed parameter to specify the time in seconds from now when the URL should expire. If specified, the URL contains the expiry timestamp in the URL, and the image signature is modified accordingly.

Examples of generating URLs

1. Chained Transformations as a query parameter

    image_url = imagekit.url({
        "path": "/default-image.jpg",
        "url_endpoint": "",
        "transformation": [{
                    "height": "300",
                    "width": "400"
                    "rotation": 90
        "transformation_position ": "query"

Sample Result URL -

2. Sharpening and contrast transforms and a progressive JPG image

There are some transforms like Sharpening that can be added to the URL with or without any other value. To use such transforms without specifying a value, specify the value as "-" in the transformation object. Otherwise, specify the value that you want to be added to this transformation.

    image_url = imagekit.url({
        "src": "",
        "transformation": [{
            "format": "jpg",
            "progressive": "true",
            "effect_sharpen": "-",
            "effect_contrast": "1"
//Note that because `src` parameter was used, the transformation string gets added as a query parameter `tr`

3. Signed URL that expires in 300 seconds with the default URL endpoint and other query parameters

    image_url = imagekit.url({
        "path": "/default-image",
        "query_parameters": {
            "p1": "123",
            "p2": "345"
        "transformation": [{
            "height": "300",
            "width": "400"
        "signed": True,
        "expire_seconds": 300

Sample Result URL,w-400/default-image.jpg?v=123&ik-t=1567358667&ik-s=f2c7cdacbe7707b71a83d49cf1c6110e3d701054

List of transformations

The complete list of transformations supported and their usage in ImageKit can be found here. The SDK gives a name to each transformation parameter, making the code simpler, making the code simpler, and readable. If a transformation is supported in ImageKit, but a name for it cannot be found in the table below, then use the transformation code from ImageKit docs as the name when using the url function.

Supported Transformation Name Translates to parameter
height h
width w
aspect_ratio ar
quality q
crop c
crop_mode cm
x x
y y
focus fo
format f
radius r
background bg
border b
rotation rt
blur bl
named n
overlay_image oi
overlay_image_aspect_ratio oiar
overlay_image_background oibg
overlay_image_border oib
overlay_image_dpr oidpr
overlay_image_quality oiq
overlay_image_cropping oic
overlay_image_trim oit
overlay_x ox
overlay_y oy
overlay_focus ofo
overlay_height oh
overlay_width ow
overlay_text ot
overlay_text_font_size ots
overlay_text_font_family otf
overlay_text_color otc
overlay_text_transparency oa
overlay_alpha oa
overlay_text_typography ott
overlay_background obg
overlay_image_trim oit
overlay_text_encoded ote
overlay_text_width otw
overlay_text_background otbg
overlay_text_padding otp
overlay_text_inner_alignment otia
overlay_radius or
progressive pr
lossless lo
trim t
metadata md
color_profile cp
default_image di
dpr dpr
effect_sharpen e-sharpen
effect_usm e-usm
effect_contrast e-contrast
effect_gray e-grayscale
original orig

File Upload

The SDK provides a simple interface using the .upload_file() method to upload files to the ImageKit Media library. It accepts all the parameters supported by the ImageKit Upload API.

The upload_file() method requires at least the file and the file_name parameter to upload a file and returns a Dict with error or success data. Use options parameter to pass other parameters supported by the ImageKit Upload API. Use the same parameter name as specified in the upload API documentation.

Simple usage

    file= "<url|base_64|binary>", # required
    file_name= "my_file_name.jpg", # required
    options= {
        "folder" : "/example-folder/",
        "tags": ["sample-tag"],
        "is_private_file": False,
        "use_unique_file_name": True,
        "response_fields": ["is_private_file", "tags"],

If the upload succeeds, error will be null, and the result will be the same as what is received from ImageKit's servers. If the upload fails, error will be the same as what is received from ImageKit's servers, and the result will be null. Learn more from the sample app in this repository.

File Management

The SDK provides a simple interface for all the media APIs mentioned here to manage your files. This also returns error and result. The error will be None if API succeeds.

1. List & Search Files

Accepts an object specifying the parameters to be used to list and search files. All parameters specified in the documentation here can be passed as it is with the correct values to get the results.

    "skip": 10,
    "limit": 10,

2. Get File Details Accepts the file ID and fetches the details as per the API documentation here


3. Get File Metadata Accepts the file ID and fetches the metadata as per the API documentation here


3. Get File Metadata from remote url Accepts the remote file url and fetches the metadata as per the API documentation here


4. Update File Details Update parameters associated with the file as per the API documentation here. The first argument to the update_field_details method is the file ID, and a second argument is an object with the parameters to be updated.

imagekit.update_file_details("file_id", {
    "tags": ["image_tag"],
    "custom_coordinates": "10,10,100, 100"

6. Delete File Delete a file as per the API documentation here. The method accepts the file ID of the file that has to be deleted.


6. Bulk File Delete by IDs Delete a file as per the API documentation here. The method accepts list of file IDs of files that has to be deleted.

imagekit.bulk_file_delete(["file_id1", "file_id2"])

6. Purge Cache Programmatically issue a cache clear request as per the API documentation here. Accepts the full URL of the file for which the cache has to be cleared.


7. Purge Cache Status

Get the purge cache request status using the request ID returned when a purge cache request gets submitted as pet the API documentation here


Utility functions

We have included the following commonly used utility functions in this package.

Authentication parameter generation

In case you are looking to implement client-side file upload, you are going to need a token, expiry timestamp , and a valid signature for that upload. The SDK provides a simple method that you can use in your code to generate these authentication parameters for you.

Note: The Private API Key should never be exposed in any client-side code. You must always generate these authentications parameters on the server-side


authentication_parameters = imagekit.get_authentication_parameters(token, expire)


    "token": "unique_token",
    "expire": "valid_expiry_timestamp",
    "signature": "generated_signature"

Both the token and expire parameters are optional. If not specified, the SDK uses the UUID to generate a random token and also generates a valid expiry timestamp internally. The value of the token and expire used to generate the signature are always returned in the response, no matter if they are provided as an input to this method or not.

Distance calculation between two pHash values

Perceptual hashing allows you to construct a has value that uniquely identifies an input image based on the contents of an image. metadata API returns the pHash value of an image in the response. You can use this value to find a duplicate or similar image by calculating the distance between the two images.

This SDK exposes phash_distance function to calculate the distance between two pHash value. It accepts two pHash hexadecimal strings and returns a numeric value indicative of the level of difference between the two images.

def calculate_distance():
    # fetch metadata of two uploaded image files
    # extract pHash strings from both: say 'first_hash' and 'second_hash'
    # calculate the distance between them:

    distance = imagekit.phash_distance(first_hash, second_hash)
    return distance

Distance calculation examples

imagekit.phash_distance('f06830ca9f1e3e90', 'f06830ca9f1e3e90')
# output: 0 (ame image)

imagekit.phash_distance('2d5ad3936d2e015b', '2d6ed293db36a4fb')
# output: 17 (similar images)

imagekit.phash_distance('a4a65595ac94518b', '7838873e791f8400')
# output: 37 (dissimilar images)

Sample Code Instruction

To run sample code go to the sample directory and run



For any feedback or to report any issues or general implementation support, please reach out to



Released under the MIT license.