Toolsky

Free Online Tools & Utilities for Every Task.

IDN/Punycode Phishing Detector

Identify domain homograph attacks using Punycode and Unicode characters.

Paste the suspicious domain here to reveal its true Unicode form.
Analysis Results
Decoded Unicode Domain (Visual Representation)

Tip: Always check the address bar in your browser. Most modern browsers will display the xn-- version if the domain is considered suspicious.

A domain name that looks perfectly familiar can still be a trap. Attackers register addresses built from look-alike characters borrowed from other alphabets, so the name on your screen reads as your bank, your email provider or a shop you trust, while the actual address points somewhere else entirely. This IDN and punycode phishing detector helps you pull back the curtain: paste in a suspicious domain and it shows you how the name is really encoded, which characters it truly contains, and whether it is mixing scripts in the way homograph attacks do.

What a homograph attack actually is

Domain names were originally limited to plain Latin letters, digits and hyphens. Internationalized Domain Names (IDNs) changed that by allowing characters from other writing systems — Cyrillic, Greek, Chinese, Arabic and many more — so people can use addresses in their own language. Behind the scenes, every IDN is converted into an ASCII form called punycode, which is why genuine international domains appear in their encoded form with an xn-- prefix. The encoding system itself is legitimate and necessary. The abuse comes from homographs: characters in different scripts that look identical or nearly identical to Latin letters. A Cyrillic "а" and a Latin "a" are different characters with different code points, but on screen they are twins. A domain assembled from such substitutes can display as a famous brand name while being a completely different, attacker-controlled address.

What this detector shows you

The tool is a focused inspection instrument rather than a general antivirus, and its job is to make the invisible parts of a domain visible:

  • It accepts a domain or full address and reveals its punycode form, including the tell-tale xn-- label that marks an internationalized name.
  • It decodes that punycode back into the real Unicode characters, so you can see what the name actually consists of.
  • It highlights script mixing — a name that combines characters from two or more writing systems is the classic signature of a homograph attack.
  • It helps you spot characters chosen because they resemble Latin letters, even when the rendered name looks flawless.

If you need to decode or encode names by hand while you investigate, the companion Punycode to Unicode converter and the reverse Unicode to Punycode tool show the raw transformation in both directions.

How to check a domain with this tool

  1. Find the suspicious address — in an email, a message, an ad or a login page — and copy the domain part exactly, without clicking through to the site first.
  2. Paste the domain into the detector's input field.
  3. Read the punycode form it reports. An xn-- prefix tells you the name contains international characters, which is worth a closer look when the visible name pretends to be an ordinary English-language brand.
  4. Study the decoded Unicode version and note which scripts the characters come from.
  5. Compare the result character by character with the genuine domain you expected. One substituted letter is enough to make it a different address.
  6. Decide how to proceed: if the scripts are mixed or a character is a look-alike, treat the link as hostile and reach the real service by typing its address yourself or using your own bookmark.

A worked example: the fake apple.com

The best-known demonstration of this attack is a domain that displays as "apple.com" but is encoded as xn--pple-43d.com. The first letter is not a Latin "a" at all — it is the Cyrillic letter а (code point U+0430). Rendered in a browser address bar, the name looks indistinguishable from Apple's real domain, yet it belongs to whoever registered it. Run that punycode label through a detector and the trick collapses: the decoded name shows a Cyrillic character sitting inside an otherwise Latin name, and the script mixing is obvious. This is exactly the kind of inspection that takes ten seconds here and is nearly impossible to do reliably with your eyes alone.

Where you are most likely to meet a lookalike domain

Homograph domains rarely appear on their own; they arrive carried by a message that wants a fast click. Phishing emails about a "suspended account" or an unexpected refund are the classic vehicle, because urgency discourages inspection. The same addresses turn up in text messages claiming a parcel is held at customs, in social-media messages promising verification badges, and in ads that imitate download buttons for popular software. Attackers also pair look-alike domains with look-alike login pages, so the address is the only clue that anything is wrong. Any time a link asks you to sign in, pay, or install something, the domain deserves a check — and if the stakes are higher, a second opinion from the Safe Browsing Checker adds another layer.

Limits of detection and the mistakes people make

An honest caveat first: an xn-- prefix is not proof of fraud. Millions of legitimate websites use internationalized names in their own languages, and a German, Japanese or Russian business with a properly spelled name in its own script is doing nothing wrong. What should raise your suspicion is the combination — an international encoding wrapped around a name that visually impersonates a brand which normally uses plain ASCII. The detector flags structure, not intent, so the final judgment is yours. The common mistakes are trusting the rendered text in an address bar, checking only the beginning of a long address (attackers pad URLs so the familiar part appears first), and assuming a padlock icon means a site is genuine — it only means the connection is encrypted, which fake sites also have. For the underlying standard, the encoding rules are defined in RFC 3492, published by the RFC Editor.

Frequently Asked Questions About Punycode and Homograph Domains

What is punycode in simple terms?

Punycode is the system that squeezes characters from any writing system into the limited set of letters, digits and hyphens that the domain name system accepts. It is how international domain names are stored and transmitted behind the scenes.

What does the xn-- prefix mean?

It marks a domain label as an encoded internationalized name. When you see xn-- at the start of a label, the visible name contains non-ASCII characters, even if it looks like plain English.

Is every xn-- domain a phishing site?

No. Most are legitimate websites using names in their own languages. It becomes suspicious when the decoded name mixes scripts or impersonates a well-known brand that has no reason to use international characters.

How is a homograph attack different from a typo domain?

A typo domain (like "goggle.com") relies on you misreading ordinary letters. A homograph domain relies on characters from other alphabets that render identically, so even careful reading of the visible text may not reveal it.

Why doesn't my browser just block these?

Modern browsers do have defences — many display the punycode form when a name mixes scripts suspiciously — but the rules vary by browser, language settings and the exact combination of characters, so some look-alikes still render in their deceptive form.

Can a homograph attack target a brand without any famous name?

Yes. Attackers also imitate smaller services — regional banks, delivery companies, university portals — precisely because people expect less polish from them and inspect less.

What should I do if a domain turns out to be a look-alike?

Do not sign in or download anything from it. Close the page, reach the real service through its address typed by hand or your own bookmark, and report the message to your email provider or IT team if you have one.

A ten-second habit before you sign in

You cannot train your eyes to separate identical twins, so do not rely on them. Whenever a link asks for credentials or payment, give its domain ten seconds in this detector first; the encoding never lies, even when the letters do.