Punycode Converter
Converts internationalised domain names to and from punycode and warns about lookalike characters.
Converts an internationalised domain between the Unicode name people read and the xn-- form DNS resolves, in either direction. The conversion is the easy half: the confusable check runs on every input, because аpple.com written with a Cyrillic а converts cleanly, registers, resolves and looks exactly right in an address bar.
Either direction. A whole URL works too — the scheme and path are stripped.
ASCII (A-label)
xn--3e0b707e.kr
What DNS actually resolves.
Unicode (U-label)
한국.kr
What a browser shows you.
Label by label
Each label is limited to 63 characters after encoding, and the whole name to 253. A short Korean or Japanese name can be well over the limit once converted, which is where registrations fail.
- 한국xn--3e0b707e12/63Encoded
- krkr2/63
Total: 15/253 characters.
What this check can and cannot catch
It catches a label that mixes scripts — a Cyrillic а among Latin letters — which is how nearly every homograph domain is built, and it names the character and its code point.
It does not catch a lookalike written entirely in one script. "раураl" in all-Cyrillic is a valid Russian-script label and mixes nothing; telling it from the real thing needs to know which brand is being imitated, which this tool has no way to know. Neither, in general, does your browser.
It also does not apply the full IDNA 2008 or UTS-46 validity rules — a name that converts cleanly here may still be rejected by a registry.
The RFC 3492 codec runs in this page. Domains are converted and checked locally — nothing is resolved, and no name you are investigating is disclosed to a DNS server by looking at it here.
How to use it
Type or paste a name in either form
한국.kr or xn--3e0b707e.kr — the direction is worked out for you. A whole URL is accepted and the scheme and path stripped.
Read the warning panel if it appears
It names the character, its code point and the Latin letter it is standing in for, and shows what the name reads as spelled in Latin. "This looks suspicious" would be advice nobody can act on.
Check the per-label length
Each label is capped at 63 characters after encoding and the whole name at 253. A short Korean or Japanese name can be well over the limit once converted, and that is where a registration fails.
Common questions
Is the domain resolved while I look at it?
No. The RFC 3492 codec runs in this page, so a name you are investigating is not disclosed to any DNS server by being typed here — which matters when the name is one you suspect.
What does this check miss?
A lookalike written entirely in one script. All-Cyrillic "раураl" mixes nothing and is a valid Russian-script label; telling it from the real thing requires knowing which brand is being imitated, which this tool cannot know. Nor, in general, can your browser.
Why does my browser show xn-- instead of the Korean name?
Browsers display the Unicode form only when the name passes their own script rules; a mixed-script label is deliberately shown as punycode so it cannot masquerade. Seeing xn-- in the address bar is the browser warning you.
Can I register anything that converts here?
Not necessarily. This applies the punycode algorithm, not the full IDNA 2008 or UTS-46 validity rules, and every registry adds its own restrictions on top. A clean conversion is a necessary condition, not a sufficient one.
Feed it from
URL Encoder & Decoder
Percent-encodes and decodes URLs and query values, with a component-versus-whole-URL mode.
URL Parser & Query Editor
Splits a URL into its parts, shows query parameters as an editable table and rebuilds the result.
Suspicious URL Check
Inspects a link’s shape before you open it — lookalike characters, IP-literal hosts, @ tricks, deep subdomains, odd ports, shorteners — and on request traces its redirects and looks up the domain’s registration date. No reputation database.
WHOIS Lookup
Shows domain registration, registrar, nameservers and expiry, plus RDAP data where it exists.
Next
These accept what this tool produces, so the result can go straight in.
Hangul Romaniser
Converts Hangul to Roman letters under the Revised Romanisation rules, with name and passport modes.
URL Encoder & Decoder
Percent-encodes and decodes URLs and query values, with a component-versus-whole-URL mode.
URL Parser & Query Editor
Splits a URL into its parts, shows query parameters as an editable table and rebuilds the result.
Suspicious URL Check
Inspects a link’s shape before you open it — lookalike characters, IP-literal hosts, @ tricks, deep subdomains, odd ports, shorteners — and on request traces its redirects and looks up the domain’s registration date. No reputation database.
DNS Lookup
Queries A, AAAA, CNAME, MX, TXT, NS, SOA, SRV, CAA and PTR records over DoH, with TTL and the DNSSEC bit shown, from Cloudflare or Google.
HTTP Request Builder
Composes and sends a GET/POST/PUT/PATCH/DELETE/HEAD request with custom headers and a body, and shows the status, headers, body and timing — without ever forwarding cookies or Authorization.