A free collection of fast, privacy-friendly tools for image editing, face detection, color correction, visual deduplication, and automatic tagging. No installs. No uploads. 100% browser-based.
Like the tools? ☕ Help keep them free and fast — 💖 Donate on Ko-fi
Automatically buckets your images into "Faces Detected" vs "No Faces Detected."
Want more flair? Enable emotion tagging and filter by 😡 angry, 😢 sad, 😊 happy, and more.
Perfect for organizing photo shoots, datasets, or just filtering selfies from screenshots.
Fix dull, off-balance, or inconsistent images with auto white balance, histogram normalization, and per-image tuning.
Drop a folder, adjust sliders, boom — beauty.
Ideal for dataset prep, batch editing, or rescuing old JPEGs.
Group together images that *look* the same — even if they're resized, compressed, or edited slightly.
Uses perceptual hashing and DCT magic to find near-duplicates.
Great for pruning training data, backups, or meme folders gone wild.
Automatically label your images with tags like 1girl, smile,
blue_eyes, and more using a WD14 ONNX model — all in your browser, no
server required.
Also a powerful tag manager: import your existing tags by just dropping in a .txt file
that matches the filename of an image. Search, sort, and edit tags live.
Whether you're prepping datasets for training or just organizing your waifu collection, this tool has you covered.
The Ultimate Image Tool Suite was created to make image processing effortless, private, and lightning-fast. Designed for digital artists, AI researchers, hobbyists, meme archivists, and photographers alike, this suite brings together four powerful tools that cover the most common image workflow needs: sorting, deduplication, enhancement, and tagging. Whether you're curating a dataset, organizing personal photos, or refining AI training inputs, these tools are crafted to optimize your workflow.
Everything runs 100% client-side using modern web technologies like JavaScript, WebAssembly, and ONNX-powered machine learning models. This means no logins, no data collection, no servers, and zero image uploads. Your content stays securely on your device at all times, offering a seamless and private user experience. This approach also ensures blazing-fast performance, since there's no need to wait for uploads, downloads, or backend processing.
Creating a clean and well-labeled image dataset is essential for many AI and computer vision tasks, from LoRA fine-tuning to style classification and facial recognition. Here's how our tools can help streamline every stage:
1girl, solo,
blue_hair, smile, and many others. It also lets you import and manage your own
tag files, edit tags in bulk, and export everything easily for training or archiving.
At the heart of the Ultimate Image Tool Suite is a commitment to privacy and performance. All tools are engineered to run locally, in your browser. This ensures full privacy — your files never leave your device — while delivering instantaneous results thanks to hardware-accelerated graphics and local memory usage.
The technology stack behind the suite includes modern frontend frameworks, advanced client-side image manipulation libraries, and ONNX.js for efficient in-browser neural network inference. The face detection and emotion recognition modules rely on deep learning models optimized for browser execution, while the deduplication tool uses perceptual hashing algorithms inspired by human vision.
No external servers are involved at any stage. This not only protects your data but also removes any dependency on third-party APIs or slow cloud services. The result: a fast, responsive, and reliable set of tools that you can use anywhere, even offline.
Yes! Every feature in the suite is completely free to use. We don’t lock essential tools behind paywalls. Donations via Ko-fi are welcome to support future development and hosting.
Absolutely. Privacy is a core principle of this project. None of your images are uploaded, logged, or analyzed remotely. The entire suite runs client-side using JavaScript, and the codebase is open for inspection if you’re technically inclined.
Some tools are optimized for desktop use due to drag-and-drop and folder selection requirements. However, simpler tasks — like tagging or reviewing small image batches — work on modern mobile browsers like Safari and Chrome.
The suite supports standard image formats including JPEG, PNG, and WebP. Support for formats like BMP, TIFF, or GIF may depend on browser capabilities. The export formats are usually the same as input, and lossless options are preserved whenever possible.
Yes. The tools are regularly tested with thousands of images. Performance scales well on modern hardware, and memory usage is optimized. You can safely process large folders with confidence — and without risking your data privacy.
This suite is ideal for creators, data scientists, AI researchers, digital artists, hobbyists, and anyone who deals with large volumes of images. Whether you’re cleaning up your waifu collection, tagging assets for Stable Diffusion training, or organizing reference material, the suite is designed to save you time and keep you in control.
The Face Detection Sorter is a privacy-respecting, browser-based tool designed to help users automatically identify and sort images based on the presence or absence of human faces. Powered by advanced facial recognition models and facial landmark detection, this tool scans your image dataset and organizes it into categories like “Face Detected” and “No Face Detected.” This is especially useful for AI dataset curation, photo library organization, and pre-processing content for machine learning.
But it doesn’t stop at simple detection — this tool also supports emotion tagging using real-time analysis of facial expressions. You can filter and categorize images based on detected emotions such as happiness, anger, sadness, or neutrality, making it a powerful resource for sentiment-aware AI training, storytelling, or emotional dataset classification. All detection is done client-side in the browser, ensuring your images are never uploaded and remain fully private.
Whether you're cleaning up selfies, sorting family albums, or building datasets for AI facial recognition models, the Face Detection Sorter gives you speed, control, and peace of mind.
The Bulk Image Color Fixer is a robust, client-side image processing tool designed to correct lighting issues, fix color balance, and normalize histograms across entire folders of images — all from your web browser. Whether you’re dealing with a batch of underexposed photos, preparing images for a print or digital publication, or refining a training dataset for deep learning, this tool makes your job fast and effortless.
It supports global and per-image controls, enabling fine-tuned adjustments like exposure correction, temperature shifts, contrast tweaks, and saturation balancing. With features like auto white balance, histogram normalization, and intuitive sliders, you can standardize the visual appearance of thousands of images in just a few minutes. This is essential for use cases involving dataset pre-processing, AI art pipelines, and photo restoration.
Everything runs 100% offline in the browser using JavaScript and WebAssembly, meaning your images stay secure and never leave your device. Check out the Bulk Image Color Fixer to bring consistency and clarity to your image workflow.
The Visual Image Deduplicator is a must-have for anyone working with large image collections. It uses sophisticated perceptual hashing (pHash) and Discrete Cosine Transform (DCT) fingerprinting algorithms to identify visually similar or identical images — even if they’ve been resized, lightly edited, watermarked, or compressed differently. Unlike basic byte-level comparisons, this tool detects actual visual similarity, making it perfect for decluttering AI datasets, archive libraries, or meme folders.
You can process thousands of images at once to find and group duplicates into buckets, then review and manually select which version to keep. Whether you’re managing training images for Stable Diffusion, cleaning up scraped or downloaded content, or removing redundant renders, this tool helps ensure a clean and optimized image set.
Best of all, the Visual Image Deduplicator runs entirely client-side for maximum privacy, making it suitable for handling sensitive or proprietary images without risk.
The Image Auto-Tagger & Tag Manager is a cutting-edge in-browser tool that uses the
WD14 ONNX model to generate highly descriptive, dataset-ready tags for any image you upload
— without ever sending files to a server. Using advanced ONNX machine learning, this tool can identify a
wide variety of visual elements, including 1girl, open_mouth,
long_hair, outdoors, and many others. It’s ideal for preparing image datasets for
training models like Stable Diffusion or DreamBooth, managing galleries,
or simply organizing your anime image folder.
Tags are sorted by confidence score, and you can set thresholds to only keep the most relevant labels. You can also import your own tags via `.txt` files (automatically matched by filename), edit them inline in the browser, and export everything neatly. It works perfectly with tagged datasets used for AI training or labeling pipelines.
If you're looking to automate your image labeling, reduce manual sorting time, or simply enhance dataset quality, the Image Auto-Tagger & Tag Manager is your go-to solution.
Imagine: You ripped a ton of images from a site. Now you want to organize the data set, or maybe prep it for a
LoRA training. You've got hundres of random images... where do you even start!?
First, take them to the Deduplicator - this can remove any images that are identical, or even slighly close to
eachother (such as different expressions).
Next up, you can use the Face Detection tool to remove remove any images that don't contain a face. You can use
basic emotion tagging here, to also sort and look for certain expressions / emotions.
After you've cleaned up your data set, you can use the color fixer to bulk edit the images to make sure the
images all look nice and crisp.
Finally, take your data set and drop them in the Auto-Tagger. Here you can run the images through a WD14 tagger
model to automatically generate tags for each image. If you already have tagged images you can drag the images
and their matching .txt files - this will load all the tags and you can search, sort, and organize
the images based on the tags. After you're done you can download a zip of the .txt tag files.