Nuke 发布版本: Nuke 7.2

A powerful image loading and caching system. It makes simple tasks like loading images into views extremely simple, while also supporting more advanced features for more demanding apps.

  • Fast LRU memory cache, native HTTP disk cache, and custom aggressive LRU disk cache.
  • Progressive image loading (progressive JPEG and WebP)
  • Resumable downloads, request deduplication, prioritization, rate limiting and more
  • Alamofire, WebP, Gifu, FLAnimatedImage extensions
  • RxNuke - RxSwift extensions
  • Automates prefetching with Preheat (deprecated in iOS 10)

Quick Start

Upgrading from the previous version? Use a Migration Guide.

More information is available in Documentation directory and a full API Reference. When you are ready to install Nuke you can follow an Installation Guide - all major package managers are supported.


Load Image into Image View

You can load an image into an image view with a single line of code:

Nuke.loadImage(with: url, into: imageView)

Nuke will automatically load image data, decompress it in the background, store image in memory cache and display it. ƒ

To learn more about the image pipeline see the dedicated section.

Nuke keeps track of each image view. When you request a new image for a view the previous outstanding request gets cancelled. The same happens automatically when the view is deallocated.

func collectionView(_ collectionView: UICollectionView, cellForItemAt indexPath: IndexPath) -> UICollectionViewCell {
    // Previous request for the image view gets cancelled. The view is
    // automatically prepared for reuse (image set to `nil`).
    Nuke.loadImage(with: url, into: cell.imageView)

Placeholders, Transitions and More

You can use an options parameter (ImageLoadingOptions) to customize the way images are loaded and displayed. You can provide a placeholder, select one of the built-in transitions or provide a custom one:

    with: url,
    options: ImageLoadingOptions(
        placeholder: UIImage(named: "placeholder"),
        transition: .fadeIn(0.33)
    into: imageView

There is also a very common scenario when the placeholder (or the failure image) needs to be displayed with a content mode different from the one used for the loaded image. ImageLoadingOptions supports precisely that:

let options = ImageLoadingOptions(
    placeholder: UIImage(named: "placeholder"),
    failureImage: UIImage(named: "failure_image"),
    contentModes = .init(
        success: .scaleAspectFill,
        failure: .center,
        placeholder: .center

Nuke.loadImage(with: url, options: options, into: imageView)

If you find yourself in a situation where you need some feature that is not implemented by ImageLoadingOptions, it might mean that you need to use ImagePipeline directly to fetch an image and then display it.

To make all image views in the app share the same behaviour modify ImageLoadingOptions.shared.

Image Requests

Each request is represented by an ImageRequest struct. A request can be created with either URL or URLRequest.

var request = ImageRequest(url: url)
// var request = ImageRequest(urlRequest: URLRequest(url: url))

// Change memory cache policy:
request.memoryCacheOptions.isWriteAllowed = false

// Update the request priority:
request.priority = .high

Nuke.loadImage(with: request, into: imageView)

Process an Image

Nuke can process images for you. The first option is to resize the image using a Request:

/// Target size is in pixels.
ImageRequest(url: url, targetSize: CGSize(width: 640, height: 320), contentMode: .aspectFill)

To perform a custom tranformation use a processed(key:closure:) method. Her's how to create a circular avatar using Toucan:

ImageRequest(url: url).process(key: "circularAvatar") {
    Toucan(image: $0).maskWithEllipse().image

All of those APIs are built on top of ImageProcessing protocol. If you'd like to you can implement your own processors that adopt it. Keep in mind that ImageProcessing also requires Equatable conformance which helps Nuke identify images in memory cache.

See Core Image Integration Guide for more info about using Core Image with Nuke

Image Pipeline

You can use ImagePipeline to load images directly without a view. ImagePipeline offers a convenience closure-based API for loading images:

let task = ImagePipeline.shared.loadImage(
    with: url,
    progress: { _, completed, total in
        print("progress updated")
    completion: { response, error in
        print("task completed")

// task.cancel()
// task.setPriority(.high)

Tasks can be used to track download progress, cancel the requests, and dynamically udpdate download priority.

Configuring Image Pipeline

ImagePipeline is initialized with a Configuration which makes it fully customizable:

let pipeline = ImagePipeline {
    $0.dataLoader = /* your data loader */
    $0.dataLoadingQueue = OperationQueue() /* your custom download queue */
    $0.imageCache = /* your image cache */
    /* etc... */

// When you're done you can make the pipeline a shared one:
ImagePipeline.shared = pipeline

Advanced Usage

Memory Cache

Default Nuke's ImagePipeline has two cache layers.

First, there is a memory cache for storing processed images ready for display. You can get a direct access to this cache:

// Configure cache
ImageCache.shared.costLimit = 1024 * 1024 * 100 // 100 MB
ImageCache.shared.countLimit = 100
ImageCache.shared.ttl = 120 // Invalidate image after 120 sec

// Read and write images
let request = ImageRequest(url: url)
ImageCache.shared[request] = image
let image = ImageCache.shared[request]

// Clear cache

HTTP Disk Cache

To store unprocessed image data Nuke uses a URLCache instance:

// Configure cache
DataLoader.sharedUrlCache.diskCapacity = 100
DataLoader.sharedUrlCache.memoryCapacity = 0

// Read and write responses
let request = ImageRequest(url: url)
let _ = DataLoader.sharedUrlCache.cachedResponse(for: request.urlRequest)
DataLoader.sharedUrlCache.removeCachedResponse(for: request.urlRequest)

// Clear cache

Aggressive Disk Cache (Experimental)

Add a completely new custom LRU disk cache which can be used for fast and reliable aggressive data caching (ignores HTTP cache control). The new cache lookups are up to 2x faster than URLCache lookups. You can enable it using pipeline's configuration:

When enabling disk cache you must provide a keyEncoder function which takes image request's url as a parameter and produces a key which can be used as a valid filename. The demo project uses sha1 to generate those keys.

$0.enableExperimentalAggressiveDiskCaching(keyEncoder: {
    guard let data = $0.cString(using: .utf8) else { return nil }
    return _nuke_sha1(data, UInt32(data.count))

The public API for disk cache and the API for using custom disk caches is going to be available in the future versions.

Existing API already allows you to use custom disk cache by implementing DataLoading protocol, but this is not the most straightforward option.

Preheat Images

Preheating (prefetching) means loading images ahead of time in anticipation of their use. Nuke provides a ImagePreheater class that does just that:

let preheater = ImagePreheater(pipeline: ImagePipeline.shared)

let requests = {
    var request = ImageRequest(url: $0)
    request.priority = .low
    return request

// User enters the screen:
preheater.startPreheating(for: requests)

// User leaves the screen:
preheater.stopPreheating(for: requests)

You can use Nuke in combination with Preheat library which automates preheating of content in UICollectionView and UITableView. On iOS 10.0 you might want to use new prefetching APIs provided by iOS instead.

Check out Performance Guide to see what else you can do to improve performance

Progressive Decoding

To use progressive image loading you need a pipeline with progressive decoding enabled:

let pipeline = ImagePipeline {
    $0.isProgressiveDecodingEnabled = true

And that's it, you can start observing images as they are produced by the pipeline. The progress handler also works as a progressive image handler:

let imageView = UIImageView()
let task = ImagePipeline.shared.loadImage(
    with: url,
    progress: { image, _, _ in
        guard let image = image else { return }
        imageView.image = image
    completion: { response, _ in
        guard let image = response?.image else { return }
        imageView.image = image

The progressive decoding only kicks in when Nuke determines that the image data does contain a progressive JPEG. The decoder scans the data and only produces a new image when it receives a full new scan (progressive JPEGs normally have around 10 scans).

See "Progressive Decoding" demo to see progressive JPEG in practice. You can also uncomment the code that blurs the first few scans of the image which makes them look a bit nicer.

Animated Images

Nuke extends UIImage with animatedImageData property. If you enable it by setting ImagePipeline.Configuration.isAnimatedImageDataEnabled to true the pipeline will start attaching original image data to the animated images (built-in decoder only supports GIFs for now).

ImageCache takes animatedImageData into account when computing the cost of cached items. ImagePipeline doesn't apply processors to the images with animated data.

There is no built-in way to render those images, there are though two integrations available: FLAnimatedImage and Gifu which are both fast and efficient.

Note: GIF is not the most efficient format for transferring and displaying animated images. The current best practice is to use short videos instead of GIFs (e.g. MP4, WebM). There is a PoC available in the demo project which uses Nuke to load, cache and dispay an MP4 video.


WebP support is provided by Nuke WebP Plugin built by Ryo Kosuge. Please follow the intructions from the repo to install it.


RxNuke adds RxSwift extensions for Nuke and enables many common use cases:

Here's an example of how easy it is to load go flow log to high resolution:

let pipeline = ImagePipeline.shared
Observable.concat(pipeline.loadImage(with: lowResUrl).orEmpty,
                  pipeline.loadImage(with: highResUtl).orEmpty)
    .subscribe(onNext: { imageView.image = $0 })
    .disposed(by: disposeBag)

Image Pipeline

Nuke's image pipeline consists of roughly five stages which can be customized using the following protocols:

Protocol Description
DataLoading Download (or return cached) image data
ImageDecoding Convert data into image objects
ImageProcessing Apply image transformations
ImageCaching Store image into memory cache

All those types come together the way you expect:

  1. ImagePipeline checks if the image is in memory cache (ImageCaching). Returns immediately if finds it.
  2. ImagePipeline uses underlying data loader (DataLoading) to fetch (or return cached) image data.
  3. When the image data is loaded it gets decoded (ImageDecoding) creating an image object.
  4. The image is then processed (ImageProcessing).
  5. ImagePipeline stores the processed image in the memory cache (ImageCaching).

Nuke is fully asynchronous (non-blocking). Each stage is executed on a separate queue tailored specifically for it. Let's dive into each of those stages.

Data Loading and Caching

A built-in DataLoader class implements DataLoading protocol and uses Foundation.URLSession to load image data. The data is cached on disk using a Foundation.URLCache instance, which by default is initialized with a memory capacity of 0 MB (Nuke stores images in memory, not image data) and a disk capacity of 150 MB.

The URLSession class natively supports the data, file, ftp, http, and https URL schemes. Image pipeline can be used with any of those schemes as well.

See Image Caching Guide to learn more about image caching

See Third Party Libraries guide to learn how to use a custom data loader or cache

Most developers either implement their own networking layer or use a third-party framework. Nuke supports both of those workflows. You can integrate your custom networking layer by implementing DataLoading protocol.

See Alamofire Plugin that implements DataLoading protocol using Alamofire framework

Memory Cache

Processed images which are ready to be displayed are stored in a fast in-memory cache (ImageCache). It uses LRU (least recently used) replacement algorithm and has a limit which prevents it from using more than ~20% of available RAM. As a good citizen, ImageCache automatically evicts images on memory warnings and removes most of the images when the application enters background.

Resumable Downloads

If the data task is terminated (either because of a failure or a cancellation) and the image was partially loaded, the next load will resume where it was left off.

Resumable downloads require server to support HTTP Range Requests. Nuke supports both validators (ETag and Last-Modified). The resumable downloads are enabled by default.

By default resumable data is stored in an efficient memory cache. Future versions might include more customization.

Request Dedupication

By default ImagePipeline combines the requests for the same image (but can be different processors) into the same task. The task's priority is set to the highest priority of registered requests and gets updated when requests are added or removed to the task. The task only gets cancelled when all the registered requests are.

Deduplication can be disabled using ImagePipeline.Configuration.


Performance is one of the key differentiating factors for Nuke. There are four key components of its performance:

Main-Thread Performance

The framework has been tuned to do very little work on the main thread. There are a number of optimizations techniques that were used to achieve that including: reducing number of allocations, reducing dynamic dispatch, backing some structs by reference typed storage to reduce ARC overhead, etc.

Robustness Under Stress

A common use case is to dynamically start and cancel requests for a collection view full of images when scrolling at a high speed. There are a number of components that ensure robustness in those kinds of scenarios:

  • ImagePipeline schedules each of its stages on a dedicated queue. Each queue limits the number of concurrent tasks. This way we don't use too much system resources at any given moment and each stage doesn't block the other. For example, if the image doesn't require processing, it doesn't go through the processing queue.
  • Under stress ImagePipeline will rate limit the requests to prevent trashing of the underlying systems (e.g. URLSession).

Memory Usage

  • Nuke tries to free memory as early as possible.
  • Memory cache uses LRU (least recently used) replacement algorithm. It has a limit which prevents it from using more than ~20% of available RAM. As a good citizen, ImageCache automatically evicts images on memory warnings and removes most of the images when the application enters background.

Performance Metrics (Beta)

When optimizing performance, it's important to measure. Nuke collects detailed performance metrics during the execution of each image task:

ImagePipeline.shared.didFinishCollectingMetrics = { task, metrics in


(lldb) po metrics

Task Information {
    Task ID - 1
    Duration - 22:35:16.123 – 22:35:16.475 (0.352s)
    Was Cancelled - false
    Is Memory Cache Hit - false
    Was Subscribed To Existing Session - false
Session Information {
    Session ID - 1
    Total Duration - 0.351s
    Was Cancelled - false
Timeline {
    22:35:16.124 – 22:35:16.475 (0.351s) - Total
    nil – nil (nil)                      - Check Disk Cache
    22:35:16.131 – 22:35:16.410 (0.278s) - Load Data
    22:35:16.410 – 22:35:16.468 (0.057s) - Decode
    22:35:16.469 – 22:35:16.474 (0.005s) - Process
Resumable Data {
    Was Resumed - nil
    Resumable Data Count - nil
    Server Confirmed Resume - nil


There are a variety extensions available for Nuke some of which are built by the community.

Name Description
RxNuke RxSwift extensions for Nuke with examples of common use cases solved by Rx
Alamofire Replace networking layer with Alamofire and combine the power of both frameworks
WebP [Community] WebP support, built by Ryo Kosuge
Gifu Use Gifu to load and display animated GIFs
FLAnimatedImage Use FLAnimatedImage to load and display animated GIFs)


  • iOS 9.0 / watchOS 2.0 / macOS 10.10 / tvOS 9.0
  • Xcode 9.2 - Xcode 10
  • Swift 4.0 - Swift 4.2


Nuke is available under the MIT license. See the LICENSE file for more info.