The Client Who Said "The Logo Is on Our Website, Just Grab It From There"
I've heard this more times than I can count. A client needs their logo on signage, on a slide deck, on a banner that's going to be 3 metres wide. They don't have a vector file — they never knew they needed one. What they have is a 120px JPG pulled from the header of a website they built in 2011. Running it through JPG-to-SVG won't give me the original designer's clean vector paths, but it gives me something I can actually work with — a scalable reconstruction that's usable for most applications and editable in Illustrator or Figma.
Upload files
Files are processed on our server and deleted automatically when the job finishes. Nothing is stored or shared.
Why "grab it from the website" isn't as simple as it sounds
A logo file embedded in a website has typically been compressed, resized, and possibly re-saved multiple times since someone first made it. A 120px JPG at that point carries all of that history in its pixel data — including JPG's compression artifacts, which appear as halos around edges and subtle color noise in what should be flat regions. That pixel history is exactly what makes JPG-to-SVG harder than PNG-to-SVG. The vectorizer has to trace through artifact noise to find the actual shapes underneath. It usually can, for simple logos — but the PNG route gives you cleaner results when you have a choice.
My actual workflow when a client only has a JPG logo
The first thing I do is convert the JPG to PNG. Not for any dramatic reason — PNG just gives the vectorizer a cleaner input. The JPG conversion to PNG step decodes the JPEG and locks the pixels losslessly, so no further compression damage accumulates. Then I run the PNG through background removal to isolate the logo on transparency. Then I convert that clean, transparent PNG to SVG.
Yes, that's three steps. But the SVG I get at the end is noticeably cleaner than going directly from the original JPG. The paths are smoother, there are fewer stray fragments, and the fill colors look right without manual correction. For a five-minute process, the improvement is significant enough that I do it every time.
If you want to go directly from JPG — and sometimes that's fine for simple logos — this converter handles it. Just know the intermediate PNG route exists if the direct result isn't clean enough.
What logo JPGs vectorize well — and which ones don't
I've vectorized enough client logos to know roughly what to expect before I start.
- Works well: Flat-color logos on a solid or white background, wordmarks with 2–4 colors, simple geometric marks, black-and-white line art. If the logo was originally designed as a vector and ended up as a JPG through some export chain, tracing usually reconstructs it recognizably.
- Works poorly: Logos with gradient fills or drop shadows — those don't have clean color boundaries for the tracer to follow. Low-quality or heavily compressed JPGs where artifact noise is visible even at normal zoom. Anything photographic or illustrative with continuous tonal variation.
The size of the source JPG matters a lot. A 40px JPG can't be traced into anything useful — there aren't enough pixels to define the shapes accurately. Use the largest version of the JPG you can find. Even if it's 300px, that's workable. Under 100px, expect rough results.
After the SVG comes out — the cleanup that makes it actually usable
The vectorized SVG usually needs a few minutes of attention before I'd hand it to a client or printer. I open it in Figma or Inkscape and do three things:
- Delete stray paths: Small isolated fragments around edges from pixel noise. Click and delete.
- Check the fill colors: Sometimes the tracer creates multiple nearly-identical color regions that should be one. Merge them and set the fill to the exact brand hex code if I have it.
- Simplify paths: The traced paths often have more points than necessary. Both Inkscape and Figma have path simplification tools that reduce point count while keeping the shape accurate.
Total time: usually under ten minutes. What I end up with is a usable vector file for a client who otherwise had nothing. It's not the same as having the original designer's clean source file — but for most practical purposes, it works.
Files are processed privately and deleted automatically
Your JPG uploads over HTTPS and is processed in temporary storage. Both the source JPG and the generated SVG are deleted when the conversion completes. No account needed.
Frequently Asked Questions
The SVG looks completely different from my logo. What happened?
Vectorization works by inferring shapes from pixel regions. If the JPG had gradients, soft shadows, or was heavily compressed, the tracer can't cleanly reconstruct the geometry. Try converting to PNG first, removing the background, then running the PNG through SVG conversion — the cleaner source produces cleaner output.
Should I convert to PNG first, then to SVG?
Often yes. PNG is lossless, so the tracer works with pixels that have no JPG artifact noise. The resulting SVG typically has cleaner, smoother paths. Use the JPG to PNG converter first if the direct result isn't good enough.
Can I use the SVG in Figma or Illustrator?
Yes. Both tools open SVG files natively. For a clean logo conversion you'll typically get a handful of editable paths. For a complex image that traced poorly, you'll get many grouped paths that are harder to work with.
Will the SVG be smaller than the JPG?
For a simple 2-color logo: often yes. For a complex image that produced thousands of paths: the SVG may be much larger than the JPG. If the SVG is unexpectedly large, the conversion probably didn't go cleanly.
Are my files stored after the conversion?
No. Files are deleted automatically as soon as the conversion job completes.
Other tools in this workflow
Explore more free tools
Need a different format or workflow? These tools are part of the same Plomz toolkit and work the same way.