What ImageMint can do
Twelve image tools, all of them free, all of them running inside your browser.
What each tool does
Every tool is its own page. Nothing here is a mode hidden inside something else, and none of them need each other to work — open the one you want and use it.
Cut something apart
One file in, many files out.
Cuts a sprite sheet into its individual frames, or splits any image into an even grid from 2×2 up to 20×20. You can drag the grid lines to line them up by eye, or let Magic mode try to find the sprite boundaries from the transparent gaps. Slices download one by one or as a ZIP.
Splits one photo into 3×3, 3×2 or 3×1 square tiles for a profile grid. You move the crop frame over the part you want, and the JPG tiles come out in a ZIP named in the order you need to post them.
Renders PDF pages to PNG or JPG at 72, 150 or 300 DPI. Take every page or a range such as 1-3, 7, and save the pages individually or as a ZIP.
Change size or shape
Same picture, different dimensions.
Resizes to an exact pixel size or a percentage of the original, one file or up to twenty at a time. Keeps the aspect ratio, can skip images already smaller than the target, and writes PNG, JPG or WebP.
Fits an image to the Instagram frame you pick — post, portrait, landscape, story, reels or profile picture — using fit, fill or stretch, and shows you what gets cropped before you download.
Scales sprites up with nearest-neighbor at whole-number factors from 2× to 32×, so every source pixel becomes a solid block and no edge softens. It also checks whether your file is a true 1× sprite or already an upscale. PNG output.
Change format or weight
Same picture, different file.
Converts between PNG, JPG and WebP, up to 20 files at once, with a quality setting and a background color to use when transparency has to be flattened into JPG.
Compresses a JPG, PNG or WebP either at a quality you choose or towards a target file size in KB, searching quality first and only reducing dimensions if that is not enough. It reports the before and after size.
Combines JPG, PNG and WebP files into a single PDF. You set the page order, pick fit-to-image, A4 or US Letter pages, and choose the margins.
Build something new
Assets assembled or altered rather than cut.
Packs a set of separate images into one sprite sheet and gives you the matching CSS, SCSS or JSON coordinates to address each frame.
Turns one image into a favicon set: a multi-size favicon.ico, an opaque apple-touch-icon, and the Android and PWA manifest PNGs, plus a true 16px preview and copy-paste head and manifest snippets.
Samples the corner pixels to work out the background color, then makes matching pixels transparent. You can adjust the tolerance and export a transparent PNG.
Where the work happens
All twelve tools use the browser's Canvas API. Your file is read by the page, worked on in memory, and written back out as a download. It is never sent to a server, because there is no server to send it to: ImageMint is a static site with no backend, no API routes, and no upload endpoint. You can check this yourself by opening your browser's network tab while you use a tool.
There is no account to create and no paid tier. Nothing is held back behind a signup.
What these tools cannot do
Knowing where a tool stops saves more time than knowing where it starts. These are the real edges.
No AI background removal
The Background Remover samples the corner pixels and removes colours close to them. That works on flat or near-flat backdrops — generated art, product shots on white, icons. It does not understand what a subject is, so a photograph with a busy or shaded background will not come out cleanly. The background removal guide shows which images work and which do not.
No text, no video
There is no OCR here: nothing reads text out of an image. There is no text-to-image generation either, and nothing on the site touches video. These tools change images you already have.
Password-protected PDFs stay closed
If you hand PDF to Image an encrypted file, it reports that the file is password protected and stops. It does not prompt for the password and has no way around one. Remove the password in the application that created the file first.
Canvas has a ceiling
Browsers cap how large a single canvas can be — roughly 16,384 pixels on a side, and about 256 megapixels in total. Output past that is refused with an explanation rather than quietly truncated, so you find out before you build a workflow on it.
Nothing is uploaded, so nothing is offloaded
Keeping your files on your machine is the whole point, and it cuts both ways. There is no rented hardware picking up the heavy part of the job: a very large image, a long PDF, or a big batch all run against whatever memory your own device has that moment. On a modern laptop that is rarely a problem; on an older phone with several tabs open, it can be.