Automated image processing for Django models
Python Shell
Pull request Compare This branch is 1 commit ahead, 516 commits behind matthewwithanm:develop.
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Failed to load latest commit information.


ImageKit is a Django app that helps you to add variations of uploaded images to your models. These variations are called "specs" and can include things like different sizes (e.g. thumbnails) and black and white versions.

For the complete documentation on the latest stable version of ImageKit, see ImageKit on RTD. Our changelog is also available.


  1. Install PIL or Pillow. If you're using an ImageField in Django, you should have already done this.
  2. pip install django-imagekit (or clone the source and put the imagekit module on your path)
  3. Add 'imagekit' to your INSTALLED_APPS list in your project's


If you've never seen Pillow before, it considers itself a more-frequently updated "friendly" fork of PIL that's compatible with setuptools. As such, it shares the same namespace as PIL does and is a drop-in replacement.

Adding Specs to a Model

Much like django.db.models.ImageField, Specs are defined as properties of a model class:

from django.db import models
from imagekit.models import ImageSpecField

class Photo(models.Model):
    original_image = models.ImageField(upload_to='photos')
    formatted_image = ImageSpecField(image_field='original_image', format='JPEG',
            options={'quality': 90})

Accessing the spec through a model instance will create the image and return an ImageFile-like object (just like with a normal django.db.models.ImageField):

photo = Photo.objects.all()[0]
photo.original_image.url # > '/media/photos/birthday.tiff'
photo.formatted_image.url # > '/media/cache/photos/birthday_formatted_image.jpeg'

Check out imagekit.models.ImageSpecField for more information.

If you only want to save the processed image (without maintaining the original), you can use a ProcessedImageField:

from django.db import models
from imagekit.models.fields import ProcessedImageField

class Photo(models.Model):
    processed_image = ImageSpecField(format='JPEG', options={'quality': 90})

See the class documentation for details.


The real power of ImageKit comes from processors. Processors take an image, do something to it, and return the result. By providing a list of processors to your spec, you can expose different versions of the original image:

from django.db import models
from imagekit.models import ImageSpecField
from imagekit.processors import ResizeToFill, Adjust

class Photo(models.Model):
    original_image = models.ImageField(upload_to='photos')
    thumbnail = ImageSpecField([Adjust(contrast=1.2, sharpness=1.1),
            ResizeToFill(50, 50)], image_field='original_image',
            format='JPEG', options={'quality': 90})

The thumbnail property will now return a cropped image:

photo = Photo.objects.all()[0]
photo.thumbnail.url # > '/media/cache/photos/birthday_thumbnail.jpeg'
photo.thumbnail.width # > 50
photo.original_image.width # > 1000

The original image is not modified; thumbnail is a new file that is the result of running the imagekit.processors.ResizeToFill processor on the original. (If you only need to save the processed image, and not the original, pass processors to a ProcessedImageField instead of an ImageSpecField.)

The imagekit.processors module contains processors for many common image manipulations, like resizing, rotating, and color adjustments. However, if they aren't up to the task, you can create your own. All you have to do is implement a process() method:

class Watermark(object):
    def process(self, image):
        # Code for adding the watermark goes here.
        return image

class Photo(models.Model):
    original_image = models.ImageField(upload_to='photos')
    watermarked_image = ImageSpecField([Watermark()], image_field='original_image',
            format='JPEG', options={'quality': 90})


ImageKit also contains a class named imagekit.admin.AdminThumbnail for displaying specs (or even regular ImageFields) in the Django admin change list. AdminThumbnail is used as a property on Django admin classes:

from django.contrib import admin
from imagekit.admin import AdminThumbnail
from .models import Photo

class PhotoAdmin(admin.ModelAdmin):
    list_display = ('__str__', 'admin_thumbnail')
    admin_thumbnail = AdminThumbnail(image_field='thumbnail'), PhotoAdmin)

AdminThumbnail can even use a custom template. For more information, see imagekit.admin.AdminThumbnail.

Image Cache Backends

Whenever you access properties like url, width and height of an ImageSpecField, its cached image is validated; whenever you save a new image to the ImageField your spec uses as a source, the spec image is invalidated. The default way to validate a cache image is to check to see if the file exists and, if not, generate a new one; the default way to invalidate the cache is to delete the image. This is a very simple and straightforward way to handle cache validation, but it has its drawbacks—for example, checking to see if the image exists means frequently hitting the storage backend.

Because of this, ImageKit allows you to define custom image cache backends. To be a valid image cache backend, a class must implement three methods: validate, invalidate, and clear (which is called when the image is no longer needed in any form, i.e. the model is deleted). Each of these methods must accept a file object, but the internals are up to you. For example, you could store the state (valid, invalid) of the cache in a database to avoid filesystem access. You can then specify your image cache backend on a per-field basis:

class Photo(models.Model):
    thumbnail = ImageSpecField(..., image_cache_backend=MyImageCacheBackend())

Or in your file if you want to use it as the default:



We love contributions! And you don't have to be an expert with the library—or even Django—to contribute either: ImageKit's processors are standalone classes that are completely separate from the more intimidating internals of Django's ORM. If you've written a processor that you think might be useful to other people, open a pull request so we can take a look!

ImageKit's image cache backends are also fairly isolated from the ImageKit guts. If you've fine-tuned one to work perfectly for a popular file storage backend, let us take a look! Maybe other people could use it.