Make your screenshots look good

Two ways to get your image in

Capture the screen here. The capture button asks your browser for a frame of any screen, window or tab you pick in its own native dialog. Nothing is installed and the frame is grabbed locally — the browser never sends it anywhere, and the permission ends the moment the capture finishes. This is desktop only; mobile browsers do not expose the screen-capture API at all, so on a phone that button is replaced by the camera and photo picker.

Bring an image you already have. Paste from the clipboard with Cmd/Ctrl+V, drag a file onto the canvas, or pick one from your device. If you already took the screenshot with your system shortcut, pasting is the fastest route — no file is written to disk at all.

Text overlay

Add a caption and drag it anywhere on the image. The text layer is real DOM while you are editing, which is why dragging is smooth and behaves identically with a mouse and with a finger; it is only rasterised into the canvas at export, at full resolution, so it stays crisp.

The background behind the text has three modes. None for text over an empty area of the frame. Solid for a hard label — the reliable choice over a busy screenshot. Translucent for the glass look, which keeps the image readable underneath while still lifting the text off it; the opacity slider is what makes that work over both light and dark screenshots.

Why a raw screenshot looks bad

A screenshot is a rectangle with hard edges that lands on whatever background the reader's app supplies. On a white slide it dissolves into the page; on a dark timeline it becomes a glaring block. There is nothing separating the content from its surroundings, so the eye has no cue about where the interface ends.

Three things fix it, and they are all this tool does. Padding creates a frame so the image is a deliberate object rather than a crop. Rounded corners read as a card, which matches how every modern interface already looks. A soft shadow lifts the image off the background, giving the depth cue that tells the eye which layer to read first. The gradient is the decorative part; the depth is what actually makes it legible.

How the rendering works

Everything is drawn on an HTML canvas. The background gradient is computed along the angle you choose, projected across the full diagonal so the transition covers the frame at any inclination rather than banding in a corner. The image is clipped to a rounded rectangle, and the shadow is cast by painting an opaque shape underneath first — drawing the shadow from the image itself would let it bleed over the artwork.

Export happens at your chosen scale, so the canvas is rendered at 2× or 3× the logical size. That is the difference between a screenshot that looks crisp when someone opens it full size and one that looks soft.

Choosing a background

The presets are pairs of neighbouring or gently complementary hues at mid lightness. That is deliberate: a background that is too dark swallows a dark-mode screenshot, and one that is too saturated competes with the interface colours inside the image. If your screenshot is mostly light, a deeper gradient gives more separation; if it is a dark dashboard, one of the lighter or neutral presets keeps the contents readable.

The custom pickers let you match a brand palette exactly, and the angle slider matters more than it looks — a diagonal gradient reads as more dynamic than a vertical one, and a vertical one is calmer for documentation.

Practical settings

Privacy

The image never leaves your computer. There is no upload, no account, no watermark and no limit on how many images you process.