Protect your brand from lookalike domain attacks.

Lookalikes from 12 scripts

The letter that fakes your domain doesn’t have to be Latin. Lookalikes are measured across these twelve writing systems, and d0ma1n checks which of them a registry would actually accept.

Latin
Latin
Кириллица
Cyrillic
Ελληνικά
Greek
العربية
Arabic
漢字
Han
한글
Hangul
カタカナ
Katakana
ひらがな
Hiragana
देवनागरी
Devanagari
ไทย
Thai
ქართული
Georgian
Հայերեն
Armenian

What makes d0ma1n different

Measured, not listed

Not a static homoglyph table. confusable-vision release 2 measures 857 lookalike pairs at the size and baseline glyphs have in running text, across every macOS system font and Roboto, calibrated against pairs with known answers. 78 are proposed as additions to Unicode’s confusables list.

Registry-aware

A lookalike only matters if someone can register it. Each one is checked against what its TLD’s registry accepts, for every delegated TLD, so you see which are real openings and which are blocked.

Font-aware

Visual similarity depends on the font. A substitution that is invisible in Arial may be obvious in Georgia. d0ma1n shows each lookalike in the font where it is most convincing.

At a glance

Inspired by dnstwist. Built on measured visual similarity instead of static tables.

Capability
Lookalike pairs857 measured (confusable-vision release 2), 78 proposed to Unicode
ScriptsLookalikes of Latin-letter domains, drawn from 12 measured scripts
Scoring0 to 1: the share of text fonts where the lookalike holds
Font awarenessMost convincing font for each lookalike
DNS resolutionA, AAAA, MX, NS with threat classification
IDNA filteringOnly PVALID characters (registrable domains)
Registry rulesEvery delegated TLD: IANA tables, ICANN rules and country-code registry policies
Reverse lookupIdentify what a suspicious domain impersonates
Runtime dependencies1 (namespace-guard)

How it works

Three open-source projects in a pipeline.

confusable-vision

Casts rays through glyph outlines in every macOS system font and Roboto, and measures which characters look alike at the size they have in running text.

→

namespace-guard

Ships the maps as runtime data. Provides skeleton(), areConfusable(), and cross-script detection.

→

d0ma1n

Generates lookalikes of your domain, checks each against its registry’s rules, scores them and resolves DNS.

Open source

Built in the open. Code MIT licensed; lookalike data CC-BY-4.0.