Loading Techie Utils…

Online Image Editor

Crop, straighten, adjust and cut out, with every step visible and reversible. Free, and runs entirely in your browser — nothing is uploaded.

About Image Editor

Tools across the top, steps down one side, the picture in the middle and the settings for whatever is in hand on the right. Crop it — free or to a ratio — straighten it, resize it, adjust it, sharpen it, warm it up, cut the background out, pad it, round its corners, drop a shadow under it. Each step is applied to the result of the last one, with the image in front of you the whole time.

The ribbon is split into Transform, Adjust, Draw, Text and Redact because those are five different jobs, and a single column showing the controls for all of them at once is always mostly showing settings for something you are not doing. Pick a tool at the top and the panel on the right becomes about that tool — "draw an arrow" and "make that arrow thick and red" are different questions, asked in different places.

Shapes go on top: rectangles square or rounded, ellipses, triangles, diamonds, pentagons, hexagons, stars, speech bubbles, lines, arrows and arrows that point both ways. Each has an outline colour, a stroke, a dash and an optional fill, so a box round something in a screenshot can be a red outline with a wash of colour inside it rather than either one or the other. There is a freehand pen beside them, and an eyedropper that takes a colour out of the picture so a callout matches something already in it.

Text is its own tab because captions have more to them than a colour and a size: ten faces, bold and italic, alignment across several lines, line height, letter spacing, and an angle. Two separate answers to the problem of text on a photograph — an outline behind the letters, which works on a quiet background, and a panel behind the whole caption, which works on a busy one. Beside them are numbered step markers that count on as you place them, and a watermark that repeats across the whole picture at an angle rather than sitting once in a corner where it can be cropped off.

Redact is the tab most people arrive looking for. Drag over an account number, a face or an address and pixelate it, blur it, block it out, or leave it alone and dim everything else so the eye goes where you want it. All four rewrite the pixels — there is no layer underneath with the original still in it, which is the whole difference between redacting a picture and covering one up.

Those marks are permanent the moment they are applied, and that is a fact about the format rather than a shortcut. An image has no layer to keep a mark in — pixels are all there is — so every mark is its own step in the history instead, and going back is one click rather than an eraser. Promising editable marks on a bitmap would be promising something it cannot hold.

The keyboard covers what the pointer is worst at, which is the last pixel of a placement. Undo and redo answer to all three spellings, the arrow keys nudge whatever is waiting to be applied — the crop rectangle, the shape, the caption — by one pixel or ten with shift, Enter applies it and Escape backs out. None of them fire while you are typing.

The rail holds the history rather than a list of files, because for one image the equivalent of "the other pages" is the states it has been through. Every step is a thumbnail with what it did and what size it left the picture, and clicking one goes back to it — undo with the destination visible, rather than pressing a button repeatedly and watching to see when it looks right.

There are already a crop tool, a resizer, a rotator and a converter here, and each of them is the right shape for what it does: one question, one answer. What none of them can be is a sequence. Cropping, then straightening, then checking whether it is still readable means three trips through three forms and three downloads, and the judgement that actually matters — does this look right — gets made on a thumbnail afterwards instead of while doing it. That is the gap this fills, and it is why those tools are still the better choice when the job really is one change.

Adjustments are shown live and written once. Dragging brightness around costs a redraw of the preview rather than a pass over the full-size image, so the sliders stay responsive on a large photograph — and nothing is baked in until Apply, so a slider moved and moved back leaves no trace.

Everything happens in this tab. The picture is never uploaded, which for photographs is not a small point: they carry faces, locations and timestamps, and most editing sites work by taking a copy of all three.

Where it helps

  • Several changes in a row, each applied to the last, with the picture visible throughout.
  • Undo, redo, and one step back to the image you opened.
  • Live preview of adjustments, applied to the full-size image only when you commit them.
  • Crop by dragging on the picture rather than by typing four numbers and hoping, free or locked to a ratio.
  • A rail of steps you can click back to, which is faster and less anxious than counting undos.
  • Fills the window when the picture deserves the whole screen, and both side panels fold away when it deserves more still.
  • Auto-trims a flat border, cuts a background out, sharpens, vignettes and frames — the things that otherwise need five separate tools.
  • Twelve shapes, a freehand pen and an eyedropper that takes its colour out of the picture itself.
  • Captions with ten faces, italic, alignment, line height, tracking and an angle, readable over anything via an outline or a panel.
  • Numbered step markers that count on as they are placed, for a screenshot that walks somebody through something.
  • A watermark tiled across the whole picture, not dropped in one corner where a crop removes it.
  • Pixelate, blur, block out or spotlight a region — redaction that rewrites the pixels rather than covering them.
  • Pads unevenly, rounds the corners and drops a shadow, which is most of what a screenshot needs before it goes in a document.
  • Shapes can be moved and resized by their handles before they are applied, and nudged a pixel at a time with the arrow keys.
  • Undo, redo, nudge, apply and cancel all have keys, and none of them fire while you are typing.
  • Warns before a format change loses transparency, and lets you choose what goes behind it.
  • Never uploads the image — it is opened, edited and written entirely in this tab.

Limits

  • It edits pixels. There are no layers and no masks, and a shape or a caption cannot be moved once applied — undo is what takes it back.
  • Warmth is a wash rather than a white-balance correction: it has no idea what in the picture was meant to be neutral, so it warms or cools everything equally.
  • Rounding the corners or adding a shadow needs transparency, so the format switches to PNG — and a PNG of a photograph is several times the size of the JPEG.
  • The history holds full-size copies of the picture, so it stops at 40 MB and sixteen steps back.
  • HEIC and SVG have to go through the Image Converter first; a canvas cannot read the one or write the other.
  • Every save re-encodes from what is on screen, so repeatedly saving as JPEG at a low quality will visibly degrade the picture.
  • Metadata is not carried over. A saved image has no EXIF, which is usually what you want and occasionally not.

Frequently asked questions

Is a blur or a pixelation safe to publish?

Pixelation is, because it averages the detail away and the information is no longer in the file. A light blur is weaker than it looks — a small radius can be reversed well enough to read a short number back — which is why the slider starts where it does. If it matters, block it out or pixelate it rather than blurring it.

Can a redaction be peeled back off afterwards?

No. All four write into the pixels, and there is no layer underneath holding the original. That is the difference between this and a black rectangle drawn over a PDF, which is the mistake that keeps making the news.

Why did my format change to PNG by itself?

Because you rounded the corners, added a shadow or cut the background out, and all three leave transparency. JPEG cannot carry it and would fill those areas with black, so the format moves to PNG rather than quietly ruining the result.

Is my photo uploaded anywhere?

No. It is read by this page, drawn onto a canvas and written back out, all inside the tab. Nothing is sent to a server, which you can confirm by opening the network panel — or by turning the network off after the page has loaded and using it anyway.

Why did my transparent background turn white?

Because JPEG cannot carry transparency, so something has to go behind it. When that applies, the editor says so before saving and lets you pick the colour. Save as PNG or WebP to keep the transparency instead.

Why is the preview smaller than my image?

It is fitted to the panel and never enlarged past life size. Everything you do is applied at full resolution regardless — the size shown at the top of the panel is the real one, and it is what gets saved.

Why is sharpening a separate button from the other adjustments?

Because the browser's filter pipeline has blur and no inverse, so sharpening is the one adjustment that cannot be previewed for free. Brightness, contrast and the rest are drawn live at preview size and cost nothing until you apply them; sharpening reads and writes every pixel, so it runs when you ask for it.

What does auto-trim actually match?

The colour of the top-left pixel, spreading inwards while the pixels still match within the tolerance you set. It is for the flat border a scanner or a screenshot leaves, not for cutting a subject out — that is what the background tool beside it does, and it works differently because the problem is different.

Can I undo after saving?

Yes. Saving writes a file out; it does not end the editing session. The history is still there, and you can keep working and save again.

Related tools

More Images & Files tools