Free Image Editor

Yes, and specifically: no account, no email, no watermark on the output, no resolution cap, no limit on how many images you edit, and no feature held back behind an upgrade. Everything on the page works on the first visit and every visit after. There is nothing to buy, because the editing runs on your own computer rather than on servers somebody has to pay for — which is the actual reason most free editors are not.

Neither. It is a web page: open it and start editing. There is nothing to download, nothing to install, and no account to create. It works the same on Windows, macOS, Linux, Android and iOS, because the only requirement is a modern browser.

No. What you save is your image with your edits and nothing added. That is worth stating plainly because a watermark on the free tier is the usual business model in this category, and it is often not mentioned until after you have done the work.

No. The file is decoded into a canvas in your browser and every edit runs there. Nothing is transmitted, nothing is stored, and once the page has loaded it keeps working with the network disconnected — which you can check by turning it off. For anything personal, that is a meaningful difference from a tool that requires an upload, and it also means there is no waiting for a large file to travel anywhere.

Because doing it in one pass is better for the image. Cropping in one tool, adjusting in another and sharpening in a third means saving a JPEG between each step, and JPEG loses a little every time it is encoded — three round trips is three lots of loss. Here the edit is held as a description of what to do and applied to the original pixels in a single pass, so the file is encoded once, at the end. The single-purpose tools are still there when one is all you need.

Yes — click Text and then click where you want it. You get font, size, weight, italic, alignment, colour, an outline, a highlight plate behind it, a drop shadow, rotation and opacity. The outline is on by default and it matters more than it sounds: white text over a photo is invisible wherever the picture behind it happens to be pale, and a thin dark outline is what fixes that. You can also add boxes, circles, lines, arrows and freehand drawing, and everything is a separate item you can drag, restyle, reorder, duplicate or delete.

Yes. Positions and font sizes are stored as fractions of the image rather than in pixels, so a caption placed halfway across a preview is halfway across the export too, at the matching size. Overlays are also drawn after everything else — after sharpening, grain and resizing — because text that has been through those picks up halos and stipple, which look like mistakes on type even when they suit the photograph underneath.

Click Hide and then click over the area, and choose Blur, Pixelate or Solid. Be aware they are not equally safe. All three destroy the detail in the exported file, so none can be undone by opening the result in an editor — but blur and pixelate are filters, and on short predictable content like a six-digit number or a plate the original can sometimes be recovered by rendering candidates and comparing. For anything that genuinely must stay private, Solid is the only one that removes the information rather than transforming it. The tool says so next to the control rather than leaving you to find out.

Yes — Add a logo or sticker places any image file as its own layer, which you can move, resize, rotate and fade. Lowering the opacity gives you a watermark that does not fight the picture. Its proportions are kept whatever the shape of the photo underneath, so a wide logo does not end up stretched on a tall image.

Yes. Press Ctrl+V anywhere on the page and a screenshot on your clipboard opens straight into the editor. When you are done, Copy image puts the result back on the clipboard so you can paste it into a chat, an issue or a document without it ever touching your downloads folder. That round trip — paste, mark up, copy — is what the annotation and redaction tools are mostly used for, and going via a file at both ends is most of the friction in it. Copying out is PNG, because that is the format the clipboard specification is built around and browsers reject the others.

Yes, for JPG and WebP. Pick a limit — 20 KB up to 1 MB — and the quality is chosen by search to land just under it, usually within a percent or two. Upload forms state their limits in kilobytes rather than in quality percentages, so naming the number beats guessing at a slider until something fits. If the limit cannot be met at the current dimensions without the image falling apart, it tells you and suggests reducing the output width instead of quietly saving something over the limit or unusably degraded.

Yes, with the Undo button or Ctrl+Z, and Redo with Ctrl+Shift+Z. Dragging a slider counts as one step rather than a hundred, so undo moves back through changes you would recognise instead of individual pixels of slider travel. Every control is also live at all times: because nothing is applied destructively, you can go back and change the exposure after cropping and sharpening without those being re-applied on top of each other.

It measures the image — the histogram, the estimated illuminant, the noise level — and proposes a correction from what it finds. It is deliberately conservative, because an automatic correction that overshoots is worse than one that under-corrects: you can see what a slider does, but you cannot easily tell that a "corrected" photo is now wrong in the other direction. Everything it sets remains adjustable afterwards.

Never. Your file on disk is not touched. The editor reads a copy into memory, and saving writes a new file alongside it. If you close the page without saving, nothing has changed anywhere.

The resize control only reduces, on purpose. Enlarging is a genuinely different problem — it needs proper resampling rather than the browser's default, which is why it has its own tool. The image upscaler does it with Lanczos resampling in linear light and lets you compare the result at 100%.

It opens anything your browser can display: JPG, PNG, WebP, GIF, AVIF and BMP. It saves PNG, JPEG or WebP. PNG and WebP keep transparency; JPEG has no alpha channel, so a transparent area is filled rather than preserved. If you are unsure, PNG is the safe choice and WebP is usually the smallest at the same quality.

The preview is rendered on a fitted copy so the controls stay responsive while you drag them, and saving re-runs the same pipeline on the full-resolution original. The settings are resolution-independent, so the result matches — but effects judged at the pixel level, sharpening and grain especially, are seen properly only at full size. Both are shown at their true scale rather than one being a guess at the other.

No imposed limit. The practical ceiling is your own machine's memory, since the image is held uncompressed while you work on it — a very large photo on a phone may struggle where a desktop will not. Nothing is capped artificially to sell you a larger allowance.