Website review
This review may include AI-generated content and is not individually verified. Domain, certificate and other facts come from the observed data on this page.
What is Shapecatcher?
Shapecatcher is a free web tool that identifies Unicode characters by shape. You draw a character in a box with your mouse, click "Recognize," and it returns the Unicode glyphs whose outlines most closely match your sketch. It is built for the common problem of knowing what a symbol looks like but not its name, code point, or how to type it.
Who it helps and how it is used
- Writers and editors who spot an unfamiliar symbol in a document and need to identify or reproduce it.
- Developers and designers looking up arrows, mathematical operators, currency signs, or typographic marks for use in text, code, or layout.
- Students and researchers encountering notation in papers who need the actual character rather than a description.
The workflow is simple: sketch the character with as many strokes as needed, optionally switch to the fill or erase tool to correct it, then recognize. If results look wrong, redrawing more precisely is the main remedy the site suggests.
Limits to keep in mind
The database covers 11,817 Unicode glyphs drawn from free fonts, and Chinese, Japanese, and Korean characters are not yet supported. A character you cannot find may simply be absent rather than badly drawn. Recognition also depends on reaching the site's recognition engine, so a "Bad Gateway" error means the service is temporarily unavailable rather than that your sketch was wrong.
For a concrete case: if you are trying to type a rare math symbol into a document, draw its distinctive strokes first—the parts that separate it from similar symbols—and compare the top results by shape before copying the character.
As a next step, if Shapecatcher returns several close matches, check the Unicode list linked from the site to confirm which block a candidate belongs to, since that narrows the choice. For LaTeX users specifically, the site points to Detexify for command lookup instead.
How do I find a Unicode character when I only know what it looks like?
Draw the character in the box on Shapecatcher and click "Recognize." The tool compares your sketch against a database of Unicode glyphs and lists the closest-looking matches, so you can identify a symbol whose name or code point you don't know.
How to get a useful match
- Draw with the mouse, adding as many strokes as the shape needs; switch to the fill or erase tool to adjust.
- Sketch the character as precisely as you can, since recognition depends on shape similarity.
- Click "Clear" to wipe the canvas and results before trying a different symbol.
- If results look poor, redraw more carefully; the page itself notes that weak performance may just mean the drawing was imprecise.
Where it falls short
The database currently holds 11,817 Unicode glyphs, drawn from free Unicode fonts to avoid licensing problems. Chinese, Japanese, and Korean characters are not yet supported, so CJK symbols won't appear. If a character isn't found, it may simply be missing from the database rather than drawn badly.
Practical scenarios
- Linguistics or math students who spot an unfamiliar symbol in a paper can sketch it and copy the matching character.
- Developers hunting for a glyph to paste into code or documentation can use it as a quick lookup when they know the shape but not the name.
- LaTeX users should look elsewhere for commands: the site points to Detexify Detexify for that purpose.
As a next step, copy the best match and verify it against an official Unicode chart before relying on it in published work.
Why can't I find Chinese, Japanese, or Korean characters on Shapecatcher?
Shapecatcher currently leaves Chinese, Japanese and Korean (CJK) characters out of its recognition database because it has not found a suitable free CJK font to use. The page explains that it only draws glyphs from free Unicode fonts to avoid licensing problems, and no adequate CJK font was available when the database was built. The recognition algorithm itself is not the limitation—the page says it could handle many more characters—so the gap is about the available font data, not the matching method.
A practical consequence: if you sketch a CJK character, the tool will still return the closest shapes it knows, but those results will be misleading because the correct character was never in the pool. Don't read a poor match as a sign your sketch was wrong.
If you need to identify a CJK character by shape, treat Shapecatcher as a tool for other scripts and try a handwriting or component-based input method instead—for example, drawing the character on a phone keyboard or using a dictionary that accepts stroke or radical input. When you return to Shapecatcher for non-CJK symbols, draw with clear, separate strokes and use the fill and erase tools to match the character's weight, since the page notes that precise drawing improves results.
What should I do if the character I drew isn't found in the results?
If a drawn character doesn't appear in Shapecatcher's results, first assume the match failed rather than that the character doesn't exist. The tool compares your sketch against a fixed database of about 11,817 Unicode glyphs, and it explicitly excludes Chinese, Japanese and Korean characters for now because no suitable free CJK font has been included. So a missing result can mean the glyph isn't in the database, or that your drawing was too rough for the recognition to rank it highly.
Practical steps
- Redraw more precisely. Use the fill and erase tools to make the strokes closer to the printed shape, then recognize again. The site itself suggests drawing as accurately as possible when results are poor.
- Check whether the character is likely covered. If it's a CJK ideograph, kana or hangul, expect no match; the site says those scripts aren't supported yet. The Unicode list on the site shows which blocks are currently included.
- Try a different route for the same goal. If you know the character's approximate name, category or a related symbol, a Unicode reference such as Unicode can help you confirm the code point. For LaTeX commands specifically, Shapecatcher points to Detexify as an alternative.
- Treat a "Bad Gateway" error separately. That message means the recognition engine is unreachable or offline, not that your character is missing. Retrying later is the sensible response.
A useful decision rule: if the shape is a non-CJK symbol and a careful redraw still fails, the glyph is probably outside the database; if it's CJK, don't spend time refining the sketch, because the limitation is coverage rather than drawing quality.
How can I get LaTeX commands for a special character I found?
Shapecatcher helps you identify a character by shape, but it does not provide LaTeX commands. Draw the symbol in its box and click "Recognize" to get the closest Unicode matches; once you have the character or its Unicode name, you can look up the corresponding LaTeX command separately. For LaTeX specifically, the site points to Detexify, which maps a hand-drawn symbol to LaTeX commands directly.
A practical workflow: use Shapecatcher when you don't even know what the character is called, then confirm the codepoint and search a LaTeX symbol reference for the command. If you already know you need LaTeX output, start with Detexify instead—it saves the intermediate step.
Two caveats from the site: Chinese, Japanese and Korean characters aren't supported yet, and recognition works best with careful, precise drawing.
What does a 'Bad Gateway' error mean when I try to recognize a drawing?
A "Bad Gateway" error on Shapecatcher means the drawing was received, but the recognition engine behind it could not be reached or is currently offline. The page itself explains this in its help section, so the problem is on the service side rather than with your sketch or browser.
What to do
- Redraw and try "Recognize" again after a short wait; a temporary outage often clears.
- Keep your drawing simple and precise so a retry is quick.
- If the error persists, report it by e-mail as the site invites, noting roughly when it happened.
What it does not mean
It is not a sign that your character is missing from the database, and it does not indicate a problem with your strokes. A character that genuinely isn't recognized usually returns weak or no matches, not a gateway failure.
For context, the tool searches a database of 11,817 Unicode glyphs and does not yet cover Chinese, Japanese or Korean characters. So if recognition returns nothing useful after the service is back, the missing coverage is the more likely explanation than the error.
User reviews (0)