Rasterise an SVG to PNG at whatever resolution you need — paste the markup or drop a file, set a size or a scale factor, and download. Because the canvas is created at the exact target dimensions rather than being scaled up afterwards, the output is crisp at any size, which is the whole point of converting from a vector in the first place. Transparency is preserved by default, or fill a background colour if you need one; choosing JPEG fills white automatically, since JPEG has no alpha and unfilled areas would otherwise come out black. There is a multi-size mode for generating a full icon set in one go. Pasted markup is sanitised before rendering, because an SVG is an executable document and running an untrusted one is a genuine security problem, not a theoretical one. Everything happens on your device.
| Use | Size | Note |
|---|---|---|
| Favicon | 32 × 32 | Also export 16 and 48 |
| App icon (iOS) | 1024 × 1024 | Store listing requires it flat, no alpha |
| App icon (Android) | 512 × 512 | Play Store listing |
| Open Graph image | 1200 × 630 | Social sharing preview |
| Retina asset | 2× the CSS size | 3× for the densest displays |
Yes, by default, for PNG and WebP. Choose JPEG and a white background is filled automatically — JPEG has no alpha channel, so transparent regions would otherwise encode as black. You can also set an explicit background colour for any format.
The viewBox is used for the aspect ratio instead, which is how most icon SVGs are written. If only one of width or height is given, the other is derived from the viewBox. Percentage dimensions are ignored, since they mean nothing without a containing element.
Rasterising can only use fonts already available to the browser; a webfont referenced from the SVG is not fetched. Convert text to paths in your vector editor before exporting and it will always match. The tool warns you when it spots a font-family in the markup.
Because it was linked to an external URL. The canvas can only draw what the SVG carries inline, and a cross-origin image also taints the canvas so encoding fails outright. Inline the image as a data URI first. The tool checks for this and tells you before converting.
Pasted markup is sanitised with DOMPurify before it is rendered, which removes scripts and event handlers, and it is loaded through a blob URL into an image rather than injected into the page. SVG is an executable format, so this matters — it is worth being wary of converters that do not mention it.
No. The conversion uses the canvas in this page. Your artwork is never transmitted, logged, or stored.
Pasted markup is sanitised with DOMPurify in SVG mode before anything else happens, which strips script elements and event-handler attributes — an SVG is an executable document, and rendering an untrusted one without sanitising it is a live XSS vector. The intrinsic size is resolved from the width and height attributes, falling back to the viewBox for the aspect ratio; percentage dimensions are ignored because they are meaningless without a containing block and drawing such an SVG at its stated size renders nothing in some browsers. The sanitised markup becomes a blob: URL that is loaded into an Image element rather than injected into the document, so nothing in it runs. That image is drawn onto a canvas sized to the exact target pixel dimensions, which is what keeps the output crisp — scaling a small raster up afterwards would not. JPEG output gets a white fill first, since JPEG has no alpha channel and unfilled transparent regions encode as black. Canvas.toBlob produces the file. External references are the one thing that cannot survive: a linked image or webfont either fails to load or taints the canvas so encoding throws, so those cases are detected in the markup and warned about up front rather than producing a silently wrong PNG.