Image Toolkit FAQ

No. Every image is decoded, resized, and re-encoded entirely inside your browser tab using the Canvas API. Nothing is sent anywhere while you process images. The only network requests this page ever makes go to Stripe, and only if you choose to subscribe.

No built-in limit. The real ceiling is whatever memory your browser and device have available for decoding and re-encoding images, not a number this tool imposes. Very large batches of very large photos will simply take longer and use more memory.

Yes. Every image is decoded to raw pixels and redrawn onto a blank canvas before being re-encoded as a new file — camera model, GPS coordinates, timestamps, and every other EXIF field live in the original file's metadata section, which a canvas simply has no way to carry over. There's no toggle for this because it isn't optional: it's a side effect of how the compression itself works, not a separate feature that could be switched off.

It's flattened to a single still frame — the first frame of the animation. Browsers can only decode one frame of a GIF onto a canvas at a time, so there's no way to preserve the animation while also compressing, resizing, or converting it. If you need the animation kept intact, don't run that file through this tool.

Because a canvas can only ever output JPEG, PNG, or WebP — it has no way to write a GIF or BMP file at all. "Keep format" maps GIF and BMP inputs to PNG, since that's the closest lossless equivalent (JPEG and WebP inputs map to themselves). It's still a genuine improvement: GIF and BMP are themselves uncompressed or poorly-compressed formats, so even the PNG fallback is usually smaller.

You'll get a clear per-file error message rather than a broken download or a stuck spinner — WebP encoding in particular isn't supported in every browser. Any other files in the same batch that don't hit the same issue still process and download normally.

Yes — the "Remove Background" tab. It runs Google's MediaPipe AI segmentation model entirely in your browser (WebAssembly, no server) to separate the subject from the background and export a transparent PNG. The first use downloads the model — around 3MB, cached afterward, so it's a one-time cost — but your photo itself is never uploaded anywhere.

It can, but it's trained and optimized for people and portraits, not arbitrary product photography the way a general-purpose tool like remove.bg aims to be. Product shots, animals, or busy/cluttered scenes may not cut out as cleanly. It's a genuine, free, fully local tool for the most common case — profile pictures and people in photos — not a universal background remover.

No. The only thing ever fetched from Google is the AI model itself (a one-time, non-personalized file download, the same category of request as loading any of this page's other library code) — that happens once and is cached by your browser. Your actual photo is decoded and processed entirely inside your browser tab and is never sent anywhere.

Just your last-chosen output format, quality, and max dimension — saved in your browser's localStorage so you don't have to reselect them every time. It never remembers the images themselves; those only ever exist in memory while a tab is open and are gone the moment you close it or refresh.

No sign-up or account with us — but processing a batch (with either "Compress & Convert" or "Remove Background") now requires an active free trial, subscription, or lifetime purchase, so you'll go through Stripe Checkout first, which does ask for a card. That's the only information you'd ever give us, and it goes straight to Stripe, not to us.

Start with a 7-day free trial (card required, nothing charged until day 8), covering both Compress & Convert and Remove Background at any batch size. After the trial it's £5/month for unlimited batches across both tools. Cancel anytime during the trial and you won't be charged at all. Prefer not to have a recurring charge? A one-time £25 lifetime purchase unlocks the same unlimited access, paid once, no trial needed, no recurring charges.

No. Stripe holds your card details but doesn't charge anything until the trial ends on day 8. Cancel any time before then from "Manage or cancel subscription" (shown once your trial is active) and you'll never be charged. We ask for a card up front because that's what makes the trial a real one — rather than a resettable browser counter, which is how the old one-free-batch model worked.

Click "Manage or cancel subscription" — it appears next to the subscribed badge once unlocked in that browser. It opens Stripe's own billing portal, where you can see your next renewal date and cancel in a couple of clicks, no email needed. It only works in the same browser you subscribed in, since there's no account here — if that's not available, email pdftoolkit@icloud.com and we'll cancel it for you.

No. Free or subscribed, every image you download comes out exactly as you configured it — no watermark, logo, or branding added anywhere.

It should, though it hasn't been formally tested on lower-end mobile hardware yet. Decoding and re-encoding images happens entirely in memory, so very large batches of very large photos are the most demanding case for Compress & Convert — a phone with limited RAM may struggle with dozens of full-resolution photos at once more than a desktop would. Remove Background adds a one-time ~3MB AI model download on first use, which is heavier over a slow mobile connection than anything else on this page.