This free fake text generator and detector shows you when/how homoglyphs, characters that look similar but are different from the characters you write with, are used.
This tool has both a fake text detector and a generator. Here is how to use them:
Fake Text (Homoglyph) Detector - Lets you identify if text you have pasted into the window has homoglyph or zero space hidden within it. Helps you identify phishing emails, spoof domains or people attempting to bypass AI detectors. Homoglyphs do not bypass Originality.ai’s accurate AI detector.
Paste text
See what characters are “fake” or “Homoglyphs”
Fake Text (Homoglyph) Generator - Let's you create homoglyph fake text in order to better understand how the process works and check if the tool your using (AI detector or similar) is resistant to homoglyph attacks.
Paste or type text
Select how much text should be replaced with homoglyph characters
Review the final text
TLDR:
Homoglyphs are characters that look like another character you are used to but are actually unique characters from a different unicode language/character set.
AI detectors used to be able to be bypassed by Homoglyphs
Homoglyph attacks can involve phishing emails and spoof domains. This free tool helps you verify if the text you are seeing with your eyes is actually what you think it is.
What are Homoglyphs
Homoglyphs are Unicode characters from different alphabets or character sets that look visually identical or extremely similar, despite having entirely different underlying codes. For instance, the Latin letter "A" and the Cyrillic letter "А" appear identical but have unique Unicode representations. Similarly, the digit "0" (zero) and the capital letter "O" look almost indistinguishable in many fonts.
Using homoglyphs, you can create text that appears perfectly normal to human readers but contains hidden substitutions invisible at first glance. This "fake text" can evade automated detection systems, such as AI-based content filters, moderation tools, or security checks—highlighting both their clever use in security testing and the potential risks they pose if misused.
Example of Common Homoglyphs:
Latin 'A' (U+0041) vs. Cyrillic 'А' (U+0410): Both appear as 'A' but are different characters.
Digit '0' (U+0030) vs. Capital 'O' (U+004F): Often indistinguishable in many fonts.
Lowercase 'l' (U+006C) vs. Digit '1' (U+0031): Can look identical, especially in sans-serif fonts.
Latin 'e' (U+0065) vs. Cyrillic 'е' (U+0435): Both resemble 'e' but are from different scripts.
Latin 'c' (U+0063) vs. Cyrillic 'с' (U+0441): Visually similar but distinct characters.
Applications of Homoglyphs
Why do we care about “homoglyphs”? What real-world harm can not being able to tell the difference between real text and fake text cause?
Text Obfuscation
Homoglyphs can be used to disguise sensitive or prohibited words, making them harder for automated systems to detect.
Example: Online platforms like a chat room use a word filter that can be susceptible to a homoglyph attack.
Attempting to Bypass AI Detectors
By substituting characters with homoglyphs, AI-generated used to be able to evade detection systems designed to identify machine-written content. This is no longer possible with accurate AI detectors. Simple AI humanizers used to use this strategy, but now AI humanizers use more advanced methods, even if they are not built to bypass detectors.
Example: This study showed replacing just 10% of characters in AI-generated text with homoglyphs led to a significant drop in detection accuracy across lower-quality AI detectors. arXiv
Phishing and Security Threats
Attackers exploit homoglyphs to create deceptive domain names and email addresses that mimic legitimate ones, tricking users into divulging sensitive information.
Example: A phishing campaign targeted Microsoft 365 users by sending emails from addresses like "Suррогt Меѕѕаցе сеոtеr," where characters were replaced with visually similar ones from other alphabets. The emails directed users to malicious sites, effectively bypassing standard email filters. Mesh | Email Security Redefined for MSPs
Our Fake Text Generator Tool
What it is: A sophisticated web-based tool that identifies and generates homoglyph characters - visually identical or similar Unicode characters from different scripts that can be used for testing, researching, and learning.
How it works: The tool operates in two modes:
Detector Mode: Analyzes input text to identify suspicious characters that look like Latin letters but are actually from other Unicode scripts (Cyrillic, Greek, Mathematical symbols, etc.)
Generator Mode: Creates test cases by substituting regular Latin characters with visually similar homoglyphs from various scripts
Key Features:
Real-time highlighting of suspicious characters with color-coded script identification
Detailed character analysis showing Unicode code points, script origins, and Latin equivalents
Interactive controls including intensity sliders, visual similarity filters, and zero-width space detection
Regeneration capabilities for testing different character substitutions
Statistical overview showing percentage of fake characters and script distribution
Educational tooltips explaining each detected character's properties
Visual Similarity Levels - The tool categorizes homoglyphs by visual similarity:
High Similarity: Cyrillic, Greek, and Mathematical characters (nearly identical)
Medium Similarity: Fullwidth and Latin Extended characters (slightly different)
Low Similarity: Armenian, Cherokee, Myanmar, Malayalam (more distinct but still confusing)
Characters Covered:
The tool supports 12 different writing systems and character sets:
Cyrillic Script
Used in Russian, Ukrainian, Bulgarian, Serbian, and other Slavic languages
Examples: а, е, о, р, х (look identical to Latin a, e, o, p, x)
Covers both uppercase and lowercase characters
Greek Alphabet
Ancient and modern Greek characters
Examples: α, ο, ρ, χ (visually similar to Latin a, o, p, x)
Particularly effective because many Greek letters are visually identical to Latin
Fullwidth Characters
Japanese/Chinese full-width variants of Latin characters
Examples: a, b, c (slightly wider versions of a, b, c)
All 26 letters (upper/lowercase) and digits 0-9 are covered
Mathematical Notation
Unicode mathematical symbols that resemble Latin letters
Examples: а (Armenian small letter a), հ (Armenian small letter h)
Less commonly known, making detection harder
Cherokee Syllabics
Indigenous North American writing system
Examples: Ꭵ (Cherokee letter i), ҽ (looks like Latin e)
International Phonetic Alphabet (IPA)
Linguistic notation symbols
Examples: ɑ (script a), ɡ (script g)
Latin Extended
Extended Latin characters with diacritics
Examples: ł (L with stroke)
CJK (Chinese, Japanese, Korean)
Ideographic characters that resemble numbers
Example: 〇 (ideographic zero, looks like 0)
Myanmar Script
Southeast Asian writing system
Example: ၀ (Myanmar digit zero)
Malayalam Script
South Indian script
Example: ഠ (resembles Latin o)
Zero-Width Characters
Invisible characters that can be inserted between text
Uses Unicode Zero-Width Space (U+200B)
Character Coverage Statistics
52 Latin letters (A-Z, a-z) have homoglyph alternatives
10 digits (0-9) are covered
Basic punctuation (period, hyphen) included
Total mappings: Over 200 individual homoglyph characters
Ethical Consideration
Homoglyphs are a powerful tool for testing system robustness, exploring the limits of AI detection, and understanding how text obfuscation works. Tools like the Fake Text Generator can be educational and valuable for developers, researchers, and educators but they must be used responsibly.
Potential Misuse
Like any technology, homoglyph substitution can be exploited. Bad actors might use fake text to:
Spread misinformation that bypasses moderation filters
Evade AI-generated content detection for academic or commercial fraud
Launch phishing campaigns using spoofed domains or emails
These actions can undermine trust and security across the internet.
Guidelines for Responsible Use
Don’t use the tool to deceive others. Avoid submitting obfuscated content to schools, publishers, or clients under false pretenses.
Use it for testing, research, or learning. Validate the robustness of AI tools, explore linguistic phenomena, or teach others about content integrity.
Stay transparent. If using homoglyphs for public or professional purposes, disclose the use where relevant.
Avoid use in spam, scams, or misinformation. These actions not only violate ethical standards they may also violate laws or platform policies.
Originality.ai provides this tool to support transparency and digital literacy. Use it to understand, not to exploit.
Conclusion
We hope this Homoglyph detector and generator helps you understand the true authenticity of written work.
The ability to drive transparency into the written word is the mission we are on and we hope this free tool helps you.
Interrested in checking out more tools? Try the originality.ai AI Humanizer.
Founder / CEO of Originality.ai I have been involved in the SEO and Content Marketing world for over a decade. My career started with a portfolio of content sites, recently I sold 2 content marketing agencies and I am the Co-Founder of MotionInvest.com, the leading place to buy and sell content websites. Through these experiences I understand what web publishers need when it comes to verifying content is original. I am not For or Against AI content, I think it has a place in everyones content strategy. However, I believe you as the publisher should be the one making the decision on when to use AI content. Our Originality checking tool has been built with serious web publishers in mind!