Round image corners

This calculator checks the width, height, and proposed radius before an image pipeline applies the effect. A valid request returns the maximum permitted radius and a clear description of the rounded-rectangle alpha mask. An invalid request fails immediately, helping design tools, export scripts, and content systems reject impossible settings before rendering begins.

Validate the radius before processing an image

Enter the image width and height as positive whole pixels, then provide the desired corner radius in pixels. The validator finds the shortest side and divides it by two to establish the largest radius that can fit without opposite corner arcs crossing. For a 1200 by 800 image, for example, the shortest side is 800 pixels and the maximum valid radius is 400 pixels. A radius from zero through 400 is accepted, including fractional values when a layout or rendering engine supports subpixel geometry. Anything larger is rejected with an input error that states the calculated limit. This check is useful before calling an image transformer because it separates a cheap geometry decision from file decoding, upload, or rendering work. It also gives browser forms and APIs one consistent rule, preventing one client from silently clamping a value while another fails. Width and height must be positive integers because they represent the actual raster dimensions, while the radius must be a finite, non-negative number.

Understand the rounded-rectangle mask

A successful result describes an alpha mask with the same width and height as the source image. The central area remains opaque. At each corner, a quarter-circle arc with the requested radius defines the boundary: pixels inside the rounded rectangle remain visible, while pixels outside those four arcs become transparent when the mask is applied. The capability describes that operation but does not upload, decode, or modify an image file. A radius of zero produces an ordinary opaque rectangle, so no corner pixels are removed. At the maximum radius, the arcs meet along the shortest dimension; a square image therefore becomes a circle, while a rectangular image becomes a capsule-like rounded rectangle. The response repeats the accepted dimensions, the radius, and the computed maximum radius alongside the human-readable mask description. This makes the output suitable for logs, validation messages, design specifications, and downstream code that needs to explain what will happen before it invokes a renderer. No interpolation method is assumed because antialiasing belongs to the image engine that eventually applies the mask.

Use one geometry rule across design and automation

Consistent radius validation prevents subtle differences between mockups, websites, mobile apps, and batch image jobs. A design system can check avatar and card specifications before publishing tokens. A media service can reject an invalid transformation request before allocating image-processing resources. A storefront can verify that user-selected rounding will work for every generated thumbnail size. Because this capability performs only deterministic arithmetic and string construction, identical inputs always return identical results; it uses no network request, random value, clock, storage, or hidden image operation. The calculation preserves the source canvas dimensions because rounding corners changes visibility at the edges rather than resizing the canvas itself. The maximum radius is based only on the shortest side, so rotating the dimensions or exchanging width and height does not change the allowed limit. The API price is $0.002 per request, and the same pure calculation is suitable for a free browser runner. Use the returned description as a plan or validation record, then pass the validated radius to the separate image renderer that performs the actual pixel transformation.