Klikr has been designed for people who, while not 'visually impaired', do not have 10/10 on both eyes.
En Français (Klikr is available in Breton, Chinese, English, French, German, Italian, Japanese, Korean, Portuguese, Spanish)
Sort files by displaying pictures and documents as icons, into folders, intuitive and fast, drag-and-drop anything, hyperlegible font available in large size.
Download the DMG installer for MacOS
Download the MSI installer for Windows:
For more advanced users: run Klikr from source!
- (ypu need to be administrator)
- get updates in 1 click
- change the JVM max memory
- ...(you are frv-savvy: create your own look and feel!)
Klikr has been designed to be very intuitive. Play with Drag & Drop! Moving files and folders around has never been easier.
Contrarily to a number of other products, Klikr does not hide your images. Klikr does not use hidden folders or whatever "Libraries"! Klikr only uses 100% transparent file system operations. Klikr never modifies a file, it only creates folders when you ask and enables you to move files from folder to folder.
Klikr features a crash-resistant undo capability: actions are stored in a file allowing to undo any action, even after a crash, even for multiple files or folder.
Klikr never deletes a file without asking you for confirmation. In Klikr, "delete" actually means moving the file into the (klikr) "trash" folder
There are 3 ways to recover a deleted or accidentally moved (moved,but you did not know where) file or folder: (1) use the undo menu item (2) Use Klikr search! (3) visit the "trash" folder using Klikr (press the top right button) (4) visit the trash folder using your favorite file manager... Only clearing the "trash" folder is final, and you will be asked for confirmation. If you move a file into a folder where there is already a file with the same name, it is renamed with a postfix. When you have duplicates, and you merge by moving files, Klikr will detect identical files with the same name and move the redundant copy into the trash folder.
All Klikr data stored on your disk is located in your home folder in a folder named '.klikr' (warning for beginners: files and folders which names start with '.' are invisible by default ;-), this includes the trash, all caches and face recognition data.
Klikr supports all major image & video file formats, as well as PDF (makes icons for PDF files).
Klikr browser window displays icons for still images, animated gifs are displayed animated, PDF documents are displayed as a icon-size image of the first page, movies are displayed as a few second of animated gif (this feature requires to have ffmpeg installed, the length of the animated gif can be changed in 'preferences').
Klikr image windows support natively jpeg, png, gif, animated gif, bmp. Then if you install additional tools (GraphicMagick and ImageMagick) you get support for all the most common image formats (with a slight performance degradation).
Klikr comes with an audio players (can load music from whole folders and import YouTube URls in drag-and-drop)
Klikr relies on your default system applications for:
- video play
- PDF display/edit
- image edition
- in general opening any file that is not an image or a sound, or a not-supported format
Klikr has 2 types of windows: "Browser" and "Image".
You can open as many windows (Browser or Image) as you want, the limit is your machine's RAM.
Uses icons for images, PDFs and movies, and buttons for everything else. Has a slide show mode.
Icons size can be changed in the 'preferences' menu or using keyboard accelerators: 'meta +' or '-'
Klikr license is MIT
Klikr comes with a few look-and-feels, and they are customisable using CSS (Cascaded Style Sheets) like web pages.
The CSS files and the icons are in src/main/resources, so you can create your own L&F!
In Klikr, you can Drag-and-Drop (almost) everything!
In a Browser window:
Drag&Drop works for icons representing images in a folder
Drag&Drop works for buttons that represent folders: you can drop a file (or folder!) on them; or drag a folder into another folder.
Drag&Drop works for buttons representing non-image files in a folder
In an Image window:
Drag&Drop enables to move the image, for example, dropping an image from a Image window to a Browser window
Type "y" to move an image into the same folder as the previous move.
Up/Parent button: will open the parent directory.
Bookmark & History menu: Bookmarks, history and Undo
Files menu : enables to create a new empty folder in the current folder, and much more...
View Menu: enables to open a new browsing window (for the current folder), and more ...
Preferences menu for preferences
Trash button: will display the content of the (Klikr specific) trash folder. If you drag a file over it, the file is moved to the trash.
Opening multiple Browser-windows is very handy to sort images from one folder to multiple destination folders: just open one browser-window per folder, then move images or files using drag-and-drop.
The 'view' menu has items to create 2 side-by-side browser windows, try it!
Can load images one after the other very fast to explore a folder (using the space bar or the left/right arrows).
Speed comes from a cache with pre-loading, play with it, you will see that preloading can be "forward" if you use the right arrow to scan one image after the other or "backward" if you use the left arrow.
The slideshow mode (type "s") has variable speed, use "w" to make is slower, "x" to make it faster.
Drag&Drop drop areas (where you can drop something) include:
Browser window: the file will be moved into the corresponding folder
Folder buttons: the file will be moved into the corresponding folder
Trash button: the file will be moved into the trash folder
Up button: the file will be moved into the parent folder
You can easily rename things (folders and files). In Image mode type "r".
Klikr remembers all settings (in a human-readable file called klikr.settings).
Klikr tells you how many files, folders and pictures a folder contains, as well as the size on disk.
Klikr displays file names and pixel sizes in the title of "Image" windows.
You can sort folders in many different ways:
- by name (alphabetically)
- by date
- by file size
- by image width
- by image height
- by image aspect ratio (most compact)
- random
- mix of aspect ratio & random
- image similarity (these experimental options require a computer with a 'good' GPU)
Klikr history remembers the folders you visited in its 'history', that you can clear at any time (and it effectively erases forever the history).
Klikr uses system defaults to open files: what happens is consistent with your OS behavior.
Klikr uses system defaults to edit files: you can start the system-configured default editor for anything, from Klikr, or an application of your choice
Klikr can open/edit a file using a user-defined application, the first time you use the menu item, you will be asked to select the target applications and then it will be remembered for the next time.
You can see the full metadata of the pictures; including EXIF, etc
Search: You can find files/images by keywords (it looks for your keywords in file names).
You can close Klikr windows with Escape (or not: see preferences).
Q: What is the smallest/weakest hardware/software configuration Klikr has been proven to run onto?
A: Currently the smallest Windows machine on which Klikr is tested is a Ιntel Alder Lake N100 (3.4GHz) with 12GB of LPDDR5 @ 4800MHz and a 256GB SSD
(A: Previously, it was a HP mini-PC, Celeron N2840 2.16GHz with 2GB of RAM, Windows 8.1 (2013), however Microsoft has disabled upgrades so that in 2025 this venerable hardware cannot be tested (if you have tips, please tell!))
Q: When I type "git clone...", it says: git not found?
A: You need to install git, it is a safe open source code management tool used by all developers.
Q: It says java not found?
A: Klikr requires java, you need to install it, look for the minimum version in the installation instructions: Klikr requires a super recent version at least 23.
Q: It says gradle not found?
A: Klikr requires gradle (for expert users: Klikr also supports jbang and the mill)
Q: Why don't you use gradlew? A: Failures have been reported
Q: Where am I?
A: Klikr navigates your storage following the directory structure. The top left button makes you go up the tree (yes, computer guys are crazy, they have upside-down trees).
Q: Ok, but where am I?
A: Klikr starts in your home folder. On mac it is /Users/yourname. Often people drop things on the desktop, it is located there: /Users/yourname/Desktop
Q: Where are my images?
A: To find where your images are on your disk, use Klikr: the "Files" menu has a "Search" item. Then use Klikr bookmarks to save the paths.
Q: Can Klikr make a slide show?
A: Yes, both the image display window and the Browser window implement slide shows.
Q: What is Klikr made of?
A: Klikr is written 100% in a computer language named "java", and uses the javafx graphic system.
Q: Why is Klikr so incredibly fast?
A: Klikr uses background worker threads to avoid slowing down the User Interface.
Q: What is a "background worker thread"?
A: It is a way to execute code on the different cores of your computer. When processing is spread other multiple cores, the job gets done faster, and it does not slow down the user interface; it reacts very fast, this is what you perceive as "speed". The more cores your machine has, the faster Klikr is.
Q: Why is Klikr so slow on large images?
A: Klikr uses java ImageIO library, which is pretty fast but on very large images (say more than 100 MegaPixel) it simply requires more RAM and computing power to be decoded, if your computer is super recent and has a lot of RAM, it should be OK, but yes, on older machines it is going to take up to several seconds... and the garbage collection will cause hickups... on the other hand, I doubt you will find another FREE$ tool that can do it faster?
Q: Why is the "sort files by aspect ratio" mode slow the first time I open a folder?
A: Because the aspect ratio of every image in the folder has to be computed, before they can be sorted. Computing the aspect ratio involves opening the file and reading metadata, this is not a lot of work but if you have many images (say more than 200) in a folder, it will take several seconds on an old machine... good news is: Klikr caches these values and the second time you open the folder it should be quite fast.
Q: A popup tells me the cache is getting pretty large, what should I do?
A: To make browsing faster Klikr uses several caches in RAM and on your disk. If you have a lot of images, the caches on disk can grow so large that your main storage could become full, which is a bad thing. For this reason, Klikr monitors its disk caches and will warn you when it gets larger than a configurable limit. Use the dedicated menu to clear the caches. If you have a lot of spare room, you can change the configuration to increase the limit, to get rid of the annoying popup when you start Klikr. Note that Klikr automatically erases items that are older than 2 days from the disk caches.
Q: Should I edit the configuration file?
A: Be careful when editing the configuration file, because it is super hard to make a config file reader 100% fool-proof. The good news is: if things go wrong, just erase the file, Klikr will create a fresh clean 'default' one, but of course you will lose preferences.
Q: Why should I edit the configuration file?
A: To look for easter eggs (there used to be quite a few).
Q: How can I edit the configuration file?
A: The file name is klikr.settings, it is located in the .klikr folder in your home folder.
Q: When I start the audio player, all (or some of) my songs are gone? A: Typically, this happens because you moved your music to a different folder. Drop that folder on the audio player drop area, the songs will be reloaded and your playlists will be updated.
Q: How can I store YouTube audio tracks? A: Paste (or drag and drop) any YouTube URL in the audio player, it will download the audio track and add it to the current playlist.
Klikr can leverage several optional opensource resources. For example: ffmpeg enables to represent movies as animated gifs.
Backup: you can backup incrementally whole file trees: faster than an OS copy -r, never deletes a file, keeps track of file content (independently of names)
Deduplication: You can find duplicated files/images (even if they have different names).
Tags: You can assign tags (text strings) to images that are saved in .properties files, one per image, and Klikr moves this metadata file with the image!
Fusk: you can obfuscate images, for example if you store personal pictures on a network drive. Be careful that if you loose the key, the content will not be recoverable!
Image-ML:
ML brings to Klikr (1) image similarity and (2) face recognition
All operations are performed over HTTP to python servers that are leveraging free opensource python. It requires to have multiple python3 stuff installed: See the dedicated README in the 'scripts' folder.
The main hurdle is that you must "manually" start the python servers, for this, copy/paste the command given in the help menu.
Image search by similarity: Cosine similarity on MobileNetV3 image embeddings. The images are resized to 224x224 and passed through MobileNetV3 (28 layers Convolutional Neural Network) which produces a feature vector with 960 components.
In a folder, you can sort images by similarity and you can click on an image and ask for the 5 most similar images in the folder.
Note that since images are first down-sampled to 224x224 pixels, "similarity" will not "work" on small details. Also note that the first step consists in computing all the feature vectors (960 floats each) and store it in a cache, so if your folder contains many images it will take a while.
Image similarity de-duplication: thanks to ML, Klikr can find 'similar images' that is: images that look the same but are NOT bit-identical. This includes images that have been rescaled, or cropped, as well as minor variants like when one has been shooting a scene with action very fast, to make sure you could capture something.
Face recognition:
Detect faces in images and recognize them. It operates in 2 steps:
- You must first TRAIN the system, for that you must provide images that contain the face of the persons to be recognized and a 'tag'.
The way to "who is who" is to sort your pictures in folders, one folder per person. Then in each folder, create an empty file named ".folder_name_is_recognition_label", the folder name will become a 'tag' or class name, and images in the folder will become a training sample (assuming a face is detected...).
-
Multiple configurations can be named and saved, and training is fully INCREMENTAL: you can add persons/classes in multiple sessions to a given configuration, without having to retrain the whole system.
-
Then you can search for faces in any image. But if an image contains multiple faces, only the first "found" will be recognized, and you have no control over which one will get out "first" of the face detection...
a) MTCNN (Multi-Task Cascaded Convolutional Networks) or 4 variants of Haars Cascades can be used to detect faces in images.
b) then classification uses KNN with cosine similarity on feature vectors extracted using FaceNet 'vggface2' image embeddings.
Recommended after training: visit the folder in .klikr that has the name of your config and browse the "prototypes" i.e. the "faces" that were stored during training: you may find weird ones that are caused by face detection false positives (sometimes things that are detected as faces are not faces). Simply delete these bogus images and the next time you load the config, the vgg19 vector file will be erased too.
How to build Klikr and Klikr installers.
You need git to get the source code.
Ideally you should use sdkman for gradle and java.
/bin/bash -c "$(curl -fsSL https://raw.githubusercontent.com/Homebrew/install/HEAD/install.sh)"
brew install sdkman
sdk install gradle 9.2.0
sdk install java 25.fx-zulu
gradle klikr
gradle run
gradle audio_player
or just:
gradle audio
You need a mac.
./jpackage_for_macos.sh
the produced file is named ./Klik-1.0.dmg
You need a PC under Windows11. (maybe windows10 works too, not tested; earlier windows versions used to work with earlier Klikr+java version, but this is now hard to test as Klikr now requires a pretty recent java)
This will give you a major asset (on top of git): the git bash shell MINGw64
MINGw64 gives you a 'nix like' behavior
git clone https://github.com/philgentric/klik.git
https://sdkman.io/install/#windows-installation
sdk install gradle 9.2.0
sdk install java 25.fx-zulu
In the git bash shell MINGw64 type:
powershell -ExecutionPolicy Bypass -File jpackage_for_windows.ps1
This will create a msi file in ./build/jpackage
To get the latest version of Klikr, if you operate from source code, just click the button in the Launcher!
... Or go in the Klikr folder and type:
git pull
this will fetch the last master source code.
If you prefer an application instead of a command line tool, install github desktop: