Image Compressor

SVG Optimiser

Drop an SVG or paste markup — get a leaner file with no editor cruft, smaller path data, and no upload.

Or paste SVG markup

Lower = smaller file. 3 is invisible at any normal viewing size.

Modern vector editors (Inkscape, Illustrator, Sketch) bake a lot of metadata into their SVG exports: RDF / Dublin Core blocks, custom namespaces, comment blocks, and float coordinates with 14 decimal places. None of that matters for the rendered image and it inflates the file by 30–70% for typical exports.

This tool runs a conservative optimisation pass in your browser. It strips comments and editor metadata, rounds numeric attributes to a configurable precision, removes empty <defs>, and collapses whitespace between elements. Geometry is never modified beyond decimal rounding — visual fidelity is preserved.

Frequently asked questions

Will the optimiser change how my SVG looks?
No. The optimisations here are visually lossless — we only strip metadata, drop empty containers, round decimals (precision is configurable), and tighten whitespace. Path geometry isn't simplified, so any glyph or icon will render identically before and after.
How much will it shrink the file?
Typical Inkscape exports shrink by 40–70%. Illustrator exports vary more — sometimes 30%, sometimes 60% depending on which Adobe blocks are included. Hand-coded SVGs may already be lean and see only a few percent reduction.
Is my SVG uploaded?
No. The optimiser uses the browser's DOMParser to parse the SVG, applies transformations to the resulting DOM tree, and serialises it back to text — all locally.
What's the difference between this and SVGO?
SVGO does more — path-data simplification, attribute reordering, transform merging — at the cost of a much larger dependency. This tool delivers the 80% of value (editor-cruft removal + numeric rounding) with no external library. For ultra-fine optimisation, run the output through SVGO afterwards.
Why doesn't decimal precision lower than 3 visibly change my SVG?
Sub-pixel precision is invisible at any normal viewing size on any current display. Precision 3 keeps the file safely accurate at extreme zoom; precision 1–2 still looks identical at the sizes most icons and logos render at.