JPG to SVG.
Drop a JPG and get a vector SVG with clean shapes. The speckles JPEG leaves around edges are merged away. Free, nothing is uploaded, no sign-up.
Reading image
How to convert a JPG to SVG
- Drop your JPG. A JPG has no transparency, so a white or coloured background becomes a shape of its own in the SVG. Logos and flat illustrations work best; photos do not.
- Watch it trace. JPEG compression leaves faint noise next to edges. The colours are grouped so that noise folds into the nearest flat colour, then the shapes are traced into curves with sharp corners kept.
- Download. You get a plain SVG with one path per colour. If the background shape is not wanted, delete it in any editor, or choose the Diagram type above, which removes a plain background for you.
Questions
Is it really free?
Yes. No account, no watermark, no download limit. The site is paid for by ads.
Are my images uploaded?
No. Conversion runs on your device, inside this page. Your image never leaves your browser, so it also works with confidential files.
What happens to the white background of my JPG?
It is traced like any other colour, so you get a white shape behind the artwork. Delete that one shape in an editor, or switch the type to Diagram, which turns a plain background transparent.
Why are there small blotches around the edges?
That is JPEG compression noise. Lower Detail or Colours merges it. A larger original or a PNG copy, if you have one, gives a cleaner trace.
Is this a real vector or a JPG inside an SVG?
A real vector. Every shape is a path you can edit, recolour and scale without blur.
I have a PNG instead.
Use the PNG to SVG converter, which also keeps transparency.
What does the match score mean?
It compares the rendered SVG with your image on colour, transparency and edges. 100% would be pixel identical. Flat artwork usually lands above 99%; photos trace into many shapes and score lower, because vectors are not made for photos.
What do the settings do?
Type picks what kind of image it is, which sets sensible defaults. Colours sets how many colours the image is split into (Auto picks them from the image and adds more where needed). Detail sets how closely curves follow the pixels: higher keeps small features, lower gives smoother shapes and smaller files.