Frequently asked questions

No, and this is the core of how the site works. When you drop an image onto any tool, the file is read by your own browser using the FileReader API and loaded into your device memory. All the processing, whether that is compressing, resizing, converting or editing, happens right there using the HTML5 Canvas API built into every modern browser. The server only sends you the page itself, the same way it sends this text. Your image never travels across the internet in either direction.

Two easy ways. First: open your browser developer tools with F12, switch to the Network tab, then run any tool on a photo. You will see the page load its scripts and styles, but no outgoing request ever carries your image data. Second, and even simpler: load a tool page, switch off your wifi completely, and use the tool. It still works, because there is nothing to send and nowhere to send it. An upload-based site fails that test instantly.

There is nothing to keep. Your file exists in your browser memory while you work on it, and the result is handed straight back to you as a download. Nothing is written to any server, so there is no cache to expire, no cleanup policy to trust, and no database of user images anywhere. When you close the tab, the image is gone from memory too. The only thing counted on the server side is an anonymous page view, the same as any website.

Safer than any upload-based tool, because the document never leaves your device, so there is no copy in transit and no copy at rest on someone else's server. Two practical tips for sensitive work: use the Blur and Pixelate tool with solid pixelation rather than a light blur when hiding text or faces, since heavy pixelation destroys the underlying pixels and cannot be reversed. And run the EXIF Metadata Viewer before sharing, because photos from phones often carry GPS coordinates, the device model and a timestamp inside the file.

More than most people expect. A typical phone JPEG embeds EXIF metadata: the exact GPS location where it was taken, the date and time down to the second, the phone or camera model, and sometimes the owner's name from the device settings. Post that photo online and anyone can read it with free software. The EXIF Metadata Viewer shows you everything your photo carries, and gives you a one-click stripped copy with all of it removed.

The site uses cookies for basic, anonymous analytics, and once ads run, ad vendors set their own cookies as described in the privacy policy. You can accept or decline in the cookie banner. What matters for this FAQ: none of that touches your images. Tracking cookies see page visits, never file contents, because your files never reach the network at all.

No. The site itself is served over HTTPS, and your image never crosses the network in the first place. Someone monitoring your connection would see only that you visited imgcruncher.com, the same as they would for any site. The image work happens entirely inside your device.

Yes, and here is the honest version of why. The site costs very little to run precisely because your images are not uploaded: there are no processing servers to pay for, no storage bills growing with every user. A small display-ad income covers hosting. That is the whole business model. No sign-up, no watermarks, no daily caps, no paid tier hiding the useful features behind a price.

No. There is no registration anywhere on the site, no email capture before download, and no feature that unlocks by logging in. Open a tool, use it, leave. The Recently Used list on the tools page is stored in your own browser's localStorage, not in any account.

There is no hard limit imposed by the site. Because your own device does the work, the practical ceiling is your device, not our rules. A 30 MB photo simply takes a few seconds longer than a 3 MB one. The real boundary is pixel dimensions rather than file size: images beyond roughly 50 megapixels can fail in some browsers because the canvas that holds the decoded image runs out of memory, especially on older phones.

The pipeline tools such as the compressor and resizer accept multi-file batches, process them one after another to keep memory stable, then offer every result individually plus a single Download All as ZIP button. The Bulk Compressor is tuned for up to 20 images per run. There is no daily quota: run as many batches as you like.

Never. Watermarking someone's image to upsell them is exactly the behavior this site was built to avoid. The only watermark on this site is the one you add yourself with the Add Watermark tool, with your own text, on purpose.

Yes. Your image is yours before and after processing; the site claims no rights over anything you make with it. Compress product photos for your shop, prepare client work, publish books. There is no license fee and no attribution requirement.

You can open JPG, PNG, WEBP, AVIF, HEIC, SVG, GIF, BMP, PDF and Base64 text. You can save to JPG, PNG, WEBP, AVIF, PDF and Base64. The gap between the two lists is deliberate and honest: HEIC, SVG, GIF and BMP are read-only inputs that convert into a modern format, because browsers can decode them but cannot encode them.

A simple rule that covers most cases: WEBP for photos and general images on your own site, because it is 25 to 35 percent smaller than JPG at the same visible quality and every current browser reads it. JPG when the file must open absolutely everywhere, including old software and picky upload forms. PNG only for logos, screenshots, and anything that needs transparency or perfectly crisp edges. Never PNG for photographs: it will be five to ten times larger than the JPG for no visible gain.

AVIF is the newest mainstream image format, typically smaller than even WEBP at the same quality. All modern browsers can display it. The catch is on the creation side: encoding AVIF in the browser currently needs a Chromium-based browser such as Chrome, Edge or Brave, so the AVIF converter will not export AVIF from Safari or Firefox. If your audience uses current browsers and you care about every kilobyte, AVIF is worth it; otherwise WEBP is the safe modern choice.

Apple switched iPhones to HEIC because it stores photos at about half the size of JPG, but Windows, Android and most websites still handle it poorly or not at all. The HEIC to JPG tool decodes the HEIC right in your browser and gives you a standard JPG that opens everywhere. It takes batches too, so you can convert a whole trip's worth of photos in one run with a ZIP download.

The tools work on still images, so an animated GIF converts from its first frame only. Keeping the animation would require a video-style encoder, which these browser tools do not include. If you need the animation preserved, you want a dedicated GIF or video tool rather than an image converter, and it is better to say that plainly than to hand you a silently frozen file.

PNG and WEBP keep transparency; so does AVIF. JPG cannot store transparency at all, which is why converting a transparent PNG to JPG asks you for a background color: something has to fill the see-through areas. If you are exporting a logo or a cutout from the Background Remover, save PNG or WEBP.

An SVG is not made of pixels; it is a set of drawing instructions, which is why it scales to any size without blurring. Converting to PNG means rasterising: choosing a concrete pixel size and drawing the instructions once at that size. That is why the SVG to PNG tool asks for a scale of 1x to 4x or an exact width. Pick the largest size you will actually need, because a PNG enlarged later loses sharpness, while the SVG never does.

Base64 turns an image file into a long block of plain text that can be pasted directly inside HTML, CSS or JSON, so the image travels inside the code with no separate file and no extra network request. Developers use it for small icons and email signatures. Fair warning built into the tool: Base64 text is about 33 percent larger than the original file, so it only makes sense for small images.

Only if you push it, and you will see it coming before you commit. JPG and WEBP compression work by discarding detail your eye barely registers; at quality 75 and above the difference is invisible on screen. Below roughly quality 50 you start to see soft edges and blocky patches in smooth areas like skies. Every tool here shows you the result and the new file size before you download, so nothing degrades silently.

Type the target into the compressor's target size field, or tap a preset chip like 100 KB, and press compress. Behind the scenes the tool runs a binary search across quality levels to find the highest quality that still fits under your number. If even the lowest sensible quality cannot reach the target, it steps the pixel dimensions down and tries again, and it always tells you what it landed on. This is the tool to use when a form says "maximum 100 KB" and keeps rejecting your photo.

Because PNG is a lossless format: it has no quality dial to turn down. The compressor can only reduce a PNG's size by shrinking its pixel dimensions, and it tells you so honestly when that happens. If the image is a photograph, the real fix is converting it to JPG or WEBP first, which routinely cuts 80 to 90 percent. If it is a logo or screenshot that must stay PNG, dimension reduction is the honest option.

For web pages and social media, 70 to 80 is the sweet spot: big savings, no visible change on screen. For photos you will print or edit again later, stay at 85 to 90. For a strict file size cap, use target size mode instead and let the tool find the quality for you. And avoid re-compressing the same JPG many times over; each save discards a little more, so keep one good original and export from that.

Resizing, almost every time, and it is not close. File size scales with pixel count: a 4000 by 3000 phone photo has four times the pixels of a 2000 by 1500 copy, and no screen you post it to will show the difference. The strongest routine is resize first, then compress: on the pipeline tools you can chain both steps and run them in one pass. A photo destined for a web page rarely needs to be wider than 2000 pixels.

Shrinking an image is safe: you are throwing away pixels you no longer need, and the result stays sharp. Enlarging is the problem direction, because the missing detail has to be invented. The Image Upscaler uses good interpolation to double dimensions smoothly, and its page says plainly what it cannot do: it will not fabricate detail that was never captured, the way heavy AI upscalers claim to. Expect clean and smooth, not miraculous.

The crop and resize tools have one-tap presets for common social sizes, like 1:1 and 4:5 for Instagram or 16:9 for YouTube. For the full current numbers, including safe zones and platform file size caps, the blog keeps tested, up-to-date guides: there is a complete image size cheat sheet covering every platform, plus deep dives for Instagram, YouTube thumbnails, Facebook and TikTok written from real tests with these tools.

Everything needed ships inside your browser already. The FileReader API reads your file into memory, the Canvas API decodes it into pixels, transforms those pixels, and re-encodes the result into JPG, PNG or WEBP, and a download link hands it back. This is the same machinery every web app uses to show images; these tools just use it to edit them too. Ten years ago you genuinely needed a server for this work. Today the server in that loop is mostly a way to collect your files.

It is the site's core idea: one upload, whole job. On the pipeline tools you can stack steps, say resize to 1080 wide, then convert to WEBP, then compress to 200 KB, and run them as one pass on one upload. The steps execute in the order of the chips, pixel work first, and the image is only re-encoded once at the end, which preserves more quality than saving between separate tools. It works on batches too.

Ready-made chains for the most common jobs, so you skip the setup entirely. Email attachment fits an image under 1 MB. Instagram post crops square and sizes to 1080. Scrub for privacy strips metadata. Website hero resizes and converts to WEBP. Compress to 100 KB does exactly what it says. Tap one, drop your image, download. You can still adjust any step after applying a preset.

Yes, genuinely offline. The site is an installable app: in Chrome or Edge look for the install icon at the right end of the address bar, on Android use Add to Home Screen, on iPhone use Share then Add to Home Screen. After your first visit the tools are cached on your device, so they load and work with no internet at all, on a flight or with the wifi down. This is only possible because the processing never needed a server in the first place.

Yes. Every tool is responsive and touch-friendly, and processing runs the same way in mobile browsers as on desktop. The practical difference is capacity: phones have less memory, so a very large batch or a huge single image that a laptop handles easily may be slow or fail on an older phone. For everyday photos, phones handle it all fine.

Chrome, Edge, Firefox and Safari from the last few years all work fully, on Windows, Mac, Linux, Android and iOS. Two honest caveats: saving to AVIF needs a Chromium browser such as Chrome or Edge, and very old browsers like Internet Explorer are not supported at all because they lack the canvas features everything is built on.

Because your device is doing real work, not waiting on a network. A 40 MB, 8000 pixel wide photo has to be decoded into hundreds of megabytes of raw pixels, transformed, and re-encoded, all in browser memory. On a recent laptop that is a few seconds; on an older phone it can be noticeably longer. If a huge file is sluggish, resize it down first, then run the rest of the job on the smaller copy.

Three quick checks solve nearly every case. First, make sure JavaScript is enabled and no aggressive ad blocker or privacy extension is blocking the page's own scripts; try the tool in a private window where extensions are usually off. Second, confirm the file is really an image the tool accepts; a file renamed to .jpg is still whatever it was underneath. Third, hard-refresh the page with Ctrl+Shift+R. If it still fails, use the contact page and say which tool, which browser, and what the file was; that is enough to reproduce most bugs.

Almost always one of two things: the quality slider went too low, or a target size was too ambitious for the pixel dimensions. Forcing a 4000 pixel wide photo under 50 KB cannot end well; there are simply not enough kilobytes to describe that many pixels. The fix is to resize down first, then compress: a 1200 pixel copy compresses to 50 KB and still looks clean. As a rule, keep quality at 70 or above unless a hard limit forces your hand.

Yes, mostly by improving what you feed it. OCR works best on sharp, straight, high-contrast text: a clean screenshot reads almost perfectly, while a photo of a page taken at an angle in dim light will struggle. Crop to just the text before extracting, make sure the selected language matches the document, and expect stylised fonts and handwriting to be hit-and-miss. The tool shows its progress while reading and you can edit the result before copying it out.

Honest answer: this tool removes plain, high-contrast backgrounds well, and struggles with busy scenes and fine detail like hair, because it works by color similarity rather than a heavy AI model. Two controls fix most halos: raise the tolerance if leftover background fringe remains, and use feather to soften the cut edge. If your photo has a cluttered background, this tool will disappoint you, and its own page says the same thing.

Check three things in order. Format: some upload forms only accept JPG and PNG and will reject WEBP or AVIF, so convert to JPG for maximum compatibility. Size: forms often have caps like 2 MB or 100 KB; use target size compression to land under the cap. Dimensions: some forms require minimums, like 320 by 320 for profile photos. The tools show the output format, size and dimensions on every result, so compare those against what the form demands.

Use the contact page or email support@imgcruncher.com. Include the tool name, your browser and device, and what you expected versus what happened; a screenshot helps enormously. This is a one-person site, which has a real upside: your report goes straight to the person who wrote the code, and I read everything. Tool requests are welcome too; several tools on the site exist because someone asked.