Copy an Invisible Character and Detect Invisible Text

Copy and paste an invisible character or blank text in one click, for empty names and messages on Discord, TikTok, and Instagram. Or paste any text below to find and remove hidden characters.

More blank text & copy options ↓

Invisible Characters Revealed

Hidden characters will show up here, highlighted like ⟨ZWSP⟩. Hover one for its name, click it for details.
Characters: 0Bytes: 0Invisible: 0

How to Use Invisible Character Viewer

  1. 1Paste or type your text into the box above.
  2. 2Every hidden character lights up in the results panel, shown as a symbol or its Unicode code point.
  3. 3Hover a highlight to see its name, or tap it for details and a copy button.
  4. 4Click “Strip Invisible Characters” to remove them, or “Replace Dashes & Smart Quotes” to swap AI-style punctuation for plain keyboard characters. Both can be undone.

What Are Invisible Characters?

Invisible characters are Unicode characters that show nothing on screen but still take up space in your text. Some are blank on purpose, for empty names and messages. Others slip in when you copy text and quietly break searches, code, and formatting.

This tool highlights every one of them in pasted text, including spaces, tabs, line breaks, zero-width characters, and control codes, and removes them in one click.

Invisible Character Legend

Every character the tool detects. Click a symbol for details, or copy it.

  • NullU+0000�
  • TabU+0009	
  • Line FeedU+000A

  • Carriage ReturnU+000D
  • SpaceU+0020 
  • Non-Breaking SpaceU+00A0 
  • Soft HyphenU+00AD­
  • Combining Grapheme JoinerU+034F͏
  • Arabic Letter MarkU+061C؜
  • Hangul Choseong FillerU+115Fᅟ
  • Hangul Jungseong FillerU+1160ᅠ
  • Ogham Space MarkU+1680 
  • Khmer Vowel Inherent AqU+17B4឴
  • Khmer Vowel Inherent AaU+17B5឵
  • Mongolian Vowel SeparatorU+180E᠎
  • En QuadU+2000 
  • Em QuadU+2001 
  • En SpaceU+2002 
  • Em SpaceU+2003 
  • Three-Per-Em SpaceU+2004 
  • Four-Per-Em SpaceU+2005 
  • Six-Per-Em SpaceU+2006 
  • Figure SpaceU+2007 
  • Punctuation SpaceU+2008 
  • Thin SpaceU+2009 
  • Hair SpaceU+200A 
  • Zero-Width SpaceU+200B​
  • Zero-Width Non-JoinerU+200C‌
  • Zero-Width JoinerU+200D‍
  • Left-to-Right MarkU+200E‎
  • Right-to-Left MarkU+200F‏
  • En DashU+2013–
  • Em DashU+2014—
  • Left Single QuoteU+2018‘
  • Right Single QuoteU+2019’
  • Left Double QuoteU+201C“
  • Right Double QuoteU+201D”
  • Horizontal EllipsisU+2026…
  • Line SeparatorU+2028

  • Paragraph SeparatorU+2029

  • Left-to-Right EmbeddingU+202A‪
  • Right-to-Left EmbeddingU+202B‫
  • Pop Directional FormattingU+202C‬
  • Left-to-Right OverrideU+202D‭
  • Right-to-Left OverrideU+202E‮
  • Narrow No-Break SpaceU+202F 
  • Medium Mathematical SpaceU+205F 
  • Word JoinerU+2060⁠
  • Function ApplicationU+2061⁡
  • Invisible TimesU+2062⁢
  • Invisible SeparatorU+2063⁣
  • Invisible PlusU+2064⁤
  • ReservedU+2065⁥
  • Left-to-Right IsolateU+2066⁦
  • Right-to-Left IsolateU+2067⁧
  • First Strong IsolateU+2068⁨
  • Pop Directional IsolateU+2069⁩
  • Inhibit Symmetric Swapping (Deprecated)U+206A
  • Activate Symmetric Swapping (Deprecated)U+206B
  • Inhibit Arabic Form Shaping (Deprecated)U+206C
  • Activate Arabic Form Shaping (Deprecated)U+206D
  • National Digit Shapes (Deprecated)U+206E
  • Nominal Digit Shapes (Deprecated)U+206F
  • Braille Pattern BlankU+2800⠀
  • Ideographic SpaceU+3000 
  • Hangul FillerU+3164ㅤ
  • Byte Order Mark (ZWNBSP)U+FEFF
  • Halfwidth Hangul FillerU+FFA0ᅠ
  • Object Replacement CharacterU+FFFC
  • Language Tag (Deprecated)U+E0001󠀁
  • Cancel TagU+E007F󠁿
  • U+XXXXOther control or invisible charactersVarious

Invisible & Special Character Guides

In-depth guides for every character this tool detects: what it is, how to type it, and where it shows up. Start with the em dash and the zero-width space, or browse all characters by category.

Copy Invisible Characters

Quick Copy

Copies a zero-width space (U+200B) to your clipboard.

Manual Selection

If quick copy is blocked, select the character in this box and copy it yourself.

Common Uses of Invisible Characters

People use invisible characters on purpose for:

  • Blank usernames & profiles: Used to create empty-looking names on social, gaming, or chat platforms.(e.g. U+00A0, U+200B, U+3000)
  • Invisible messages: Send “blank” messages in apps that block normal spaces.(e.g. U+00A0)
  • Precise spacing & formatting: Fine-tune text layout in design, code, or documents with special spaces or zero-width characters.(e.g. U+2009, U+200B)
  • Bypassing filters: Slip past simple text restrictions or validation by using non-standard spaces or invisible characters.

Platform-Specific Uses

Among Us

Invisible player names

U+3164
Fortnite & PUBG

Hidden usernames

U+200B
WhatsApp

Empty messages

U+200E
Instagram

Empty comments

U+200B
YouTube

Blank video titles

U+200C
TikTok

Invisible nicknames

U+FFFC
Twitter/X

Empty tweets

U+3164

Why Detection Matters

Hidden characters break search, copy and paste, and code. Spot them before they cause problems, whether they got there by accident or on purpose.

Security Implications

  • Can obfuscate malicious content or bypass security filters.
  • Used in phishing to manipulate domain names or URLs.
  • Enable code injection in certain contexts.
  • Can fingerprint leaked documents: each copy gets its own pattern of zero-width characters.
  • Hidden prompt injection attacks on AI systems using invisible Unicode.

If you build apps, validate and clean all text input.

Does ChatGPT Use Hidden Characters?

Sometimes. AI models mostly produce visible typographic characters (em dashes, smart quotes, ellipses) that a plain keyboard doesn't type. Occasionally they also emit genuinely invisible ones: in April 2025, OpenAI's o3 and o4-mini were found inserting narrow no-break spaces (U+202F) into ordinary text.

Common AI-Injected Characters

  • Em Dash (—) U+2014: The long dash ChatGPT frequently uses to break up sentences or add dramatic pauses. Can cause issues in code or CSV files.
  • Smart Quotes (“ ” ‘ ’) U+201C U+201D U+2018 U+2019: Curly quotation marks that can break code snippets and Markdown.
  • Zero-Width Space U+200B: Completely invisible but can break string matching and URLs.
  • Narrow No-Break Space U+202F: Looks like a normal space but isn't one, so searches, comparisons, and word counts miss it.
  • Horizontal Ellipsis (…) U+2026: A single character that looks like three dots.

Does ChatGPT Watermark Text?

Yes, but not with hidden characters. The EU AI Act's transparency rules, which apply from August 2, 2026, require AI providers to mark generated text so machines can detect it. OpenAI is rolling out a watermark for ChatGPT and Codex text in the EU, and Anthropic is adding one to Claude's text worldwide. Both are statistical watermarks: a secret key nudges which words the model picks, leaving a pattern only the provider's detector can read. Nothing is added to the text, so there is nothing for this tool, or any character scanner, to find or strip.

The 2025 narrow no-break space reports were something else. OpenAI told the researchers those characters were not a watermark but “a quirk of large-scale reinforcement learning.” A watermark made of invisible characters would be useless anyway, since anyone could delete it.

What Is SynthID?

SynthID-Text is Google DeepMind's text watermarking method, published in Nature in 2024 and used in Gemini. Claude's watermark is a version of SynthID-Text, while OpenAI uses its own method called textGrain. These watermarks weaken with editing or paraphrasing and are hard to detect in short or highly factual passages. There is no public tool to check text for them yet. Detection is limited to approved researchers, regulators, and organizations, and a missing watermark doesn't prove a human wrote the text.

Why Do LLMs Use These Characters?

  • Training Data Bias: LLMs train on professionally edited texts where em dashes and smart quotes are standard.
  • Formal Tone: These characters help achieve a more polished, authoritative tone.
  • No Keyboard Limitation: Unlike humans, LLMs don't use keyboards, so typing special characters requires no extra effort.

Pro Tip

Check AI-written text here before you paste it into code, data files, or anything that needs an exact match.

Can You Detect AI-Generated Text?

No, not reliably. Despite marketing claims, AI content detection tools simply don't work well enough to trust. Even OpenAI, the makers of ChatGPT, have stated: “Do AI detectors work? In short, no, not in our experience. Our research into detectors didn't show them to be reliable enough.”

AI detectors often flag human writing as AI, even passages from the Bible or classic literature, and miss AI text that has been lightly edited.

Google doesn't penalize content just for being AI-generated. Its guidance is about “rewarding high-quality content, however it is produced,” though mass-produced pages made to game rankings still count as spam.

Adding or removing invisible characters won't “beat” a detector either. For more, see this guide to AI content detection and whether it matters.

Frequently Asked Questions

What is an invisible character?

An invisible character is a Unicode character that shows nothing on screen (or looks like a plain space) but is still in the text. Common examples are the zero-width space (U+200B), the non-breaking space (U+00A0), and control characters.

How do I copy an invisible character?

Click Copy Invisible Character at the top of this page, or the copy icon next to any character in the legend. If your browser blocks copying, select the character in the manual selection box instead.

How do I make invisible text?

Copy an invisible character, such as the zero-width space (U+200B), with the Copy Invisible Character button, then paste it where you need blank text: a username, bio, message, or caption. For longer blank text, paste it several times.

Why use invisible text?

For blank usernames and profiles, "empty" messages on apps that block normal spaces, invisible spacing in bios and captions, or watermarking. It's common on Discord, TikTok, Instagram, WhatsApp, and many games, but each platform filters characters differently, so results vary.

Can invisible characters bypass AI detectors?

Not reliably. Research has shown that swapping in look-alike (homoglyph) characters can confuse certain detectors, but invisible characters are easy to strip, many tools normalize text first, and the wording itself is unchanged. In fact, AI detectors themselves are unreliable and shouldn't be trusted.

Why are there invisible characters in my text?

They usually come along when you copy text from websites, PDFs, Word documents, or AI tools. Some are added on purpose, for formatting or to slip past filters.

Are invisible characters dangerous?

They can be. Attackers use them to hide malicious code, slip past security filters, disguise phishing URLs, and smuggle hidden prompts into AI systems.

How do I remove invisible characters?

Paste your text into the viewer above and click "Strip Invisible Characters". Everything runs in your browser, so your text never leaves your device.