Invisible Text Generator: The Complete Guide (2026)

What invisible text actually is

An invisible character is a Unicode codepoint that renders with zero width or with an empty glyph. It exists in the underlying string. So input validators see content. But the font paints nothing visible. The result is an input field that looks blank to humans while passing the platform's "must not be empty" check.

The technical category splits in two. Zero-width characters render exactly zero pixels wide. They are formatting controls. The Unicode standard reserves them for ligature control, soft hyphenation, and combining-mark separation. Empty-glyph characters are full-width characters whose visual form happens to be blank. The most famous example is Hangul Filler (U+3164), a Korean placeholder character that is treated as a normal letter by most software but has no painted shape.

Both categories serve the same cosmetic purpose. They differ in how aggressively platforms strip them. Zero-width characters are flagged by Unicode normalization as removable. Empty-glyph characters are not. Hangul Filler is therefore the most reliable invisible character in practice, even though it is technically wider than a Zero-Width Space.

The Unicode catalogue of characters that look blank

U+3164 Hangul Filler. The single most important character on this page. Officially a Korean Jamo placeholder. Renders as blank but counts as a printable letter. Survives almost every input sanitizer because it is not a control character, a whitespace character, or a formatting character. The gold standard for empty bios.

U+200B Zero-Width Space. The textbook invisible character. Width zero. No glyph. Used legitimately to mark soft word-break points in long strings. Stripped by NFC normalization and by most input trim functions. Works in many places but fails in Slack, LinkedIn, and several CMS platforms.

U+2800 Braille Pattern Blank. A Braille glyph with no raised dots. Renders as a single blank cell. Width is full. Survives most sanitizers. Note that screen readers may announce it as "blank" rather than skip silently, which is one of the few cases where the visual blank is audible.

U+034F Combining Grapheme Joiner. A pure zero-width character meant to glue graphemes for sorting purposes (used in Hebrew and some Indic scripts). Rarely stripped because it is rarely encountered by sanitizer authors. Slightly less reliable than Hangul Filler but more reliable than ZWSP.

U+2003 Em Space. A wide blank. Visually blank rather than zero-width. The fallback when Unicode-aware filters block all of the above. Counts as whitespace, so the platforms that strip whitespace will also strip this. It works in cases where filters only check for ASCII space (0x20).

U+FFA0 Halfwidth Hangul Filler. The half-width sibling of U+3164. Useful when the field has a strict character-count limit and Hangul Filler is one cell too wide. Less reliable because some platforms specifically target half-width Hangul characters during sanitation.

Other characters sometimes recommended online (Mongolian Vowel Separator U+180E, the various Variation Selectors) have either been deprecated for cosmetic use or were never reliable across platforms. Stick to the six above.

How platforms handle them in practice

The practical reliability matters more than the Unicode theory. Here is what actually happens in 2026.

Instagram. Allows Hangul Filler in bios, captions, and display names. Rejects pure ASCII whitespace. The blank Instagram bio trick uses one or two U+3164 characters.

Twitter and X. Allows Hangul Filler and Braille Pattern Blank in display names. The handle (the @username) is restricted to ASCII alphanumerics, so invisible text does not work there. Bio fields accept the full set.

WhatsApp. Allows invisible characters in status, group names, and messages. The "blank message" effect is achieved by sending a single Hangul Filler. Note that some Android keyboards strip the character before send, so paste from clipboard rather than typing.

Discord. Allows Hangul Filler and Braille Pattern Blank in usernames, server nicknames, custom status, and messages. Discord's recent username migration to the unique-handle system enforces ASCII for the handle itself but allows invisible characters in display name.

Slack. Strips most invisible characters during input normalization. Use is not reliable. Slack treats display names and channel names with aggressive trim and NFC normalization.

Telegram. Allows invisible characters in messages and group descriptions. Normalizes them away in usernames and channel handles.

Among Us and Fortnite. Both games allow Hangul Filler in display names. The "no name" trick that lets players appear as anonymous in chat. Test in a private lobby before relying on it in ranked play.

The legitimate uses behind the trick

Invisible characters exist for real reasons. The cosmetic abuse is a side effect.

Hebrew niqqud. Hebrew vowel marks attach to consonant letters using combining diacritics. Several of those diacritics are zero-width by design. Without them, vowel-pointed Hebrew (used in liturgical texts and children's books) would not render correctly.

Khmer subscripts. Khmer stacks subscript consonants below base consonants using a zero-width "coeng" marker. Removing zero-width characters from input would break Khmer text entry entirely.

Indic vowel signs. Devanagari, Tamil, Telugu, and Kannada use combining marks (matras) that attach to base consonants. Some matras are zero-width. Modern Indic input methods rely on these characters surviving Unicode normalization.

Arabic ligature control. Zero-Width Joiner (U+200D) and Zero-Width Non-Joiner (U+200C) tell the Arabic shaper whether to form a ligature at a given position. They are required for correct Arabic display in any text longer than a single word.

Line-break control. Zero-Width Space is also used legitimately to suggest a soft break in long URLs, hyphenated compounds, or CJK text where no other break opportunity exists. Web typographers insert it manually to control wrapping.

The cosmetic uses (the reason most people are here)

Empty Instagram bio. The single largest reason this generator exists. Paste one Hangul Filler character. The bio looks unset to visitors but Instagram considers it set, so the "complete your profile" nudge disappears.

Blank WhatsApp messages and status. Send a Hangul Filler as a message to display a fully blank bubble. The recipient sees an empty message with the timestamp. Useful for testing notifications, signaling "I have nothing to say but acknowledge you," or playing tricks on group chats.

Ghost-mode usernames. Gaming platforms (Among Us, Fortnite, Call of Duty Mobile) that require a non-empty display name. Hangul Filler lets a player appear nameless in chat and on the scoreboard.

Empty profile names on social. Twitter, TikTok, Snapchat, and Telegram display name fields all accept invisible characters. The aesthetic is deliberate. A profile with no visible name reads as either mysterious or finished depending on context.

Discord blank status and embeds. Discord status fields and bot embed fields commonly use invisible characters to create blank lines, spacers, or empty fields that would otherwise be rejected as empty.

The abuse vector and platform responses

Invisible text is not all harmless. Two abuse patterns have driven platform restrictions over the years.

Homograph attacks. A username that looks identical to a target user, achieved by inserting zero-width characters that the eye does not see but the database treats as distinct. The classic example is an admin impersonation in a Discord server where the impersonator's username contains one U+200B in the middle. To users, the names are visually identical. To Discord, they are different strings, so the impersonator gets the same display.

Fake-empty captchas and bypass attacks. Some older bot defenses rejected literal empty submissions but accepted any non-empty string. Attackers used invisible characters to satisfy the non-empty check while still submitting effectively nothing. Modern captcha systems normalize Unicode before checking, which closes this vector.

Platform responses have been narrow rather than blanket. Discord and Twitter restrict invisible characters in unique identifiers (handles, IDs) but allow them in display fields. Slack and Google strip them everywhere. The trade-off is between user expression and abuse prevention; each platform has made its own call.

What screen readers do with invisible characters

Screen-reader behavior varies by character and configuration.

Zero-Width Space. NVDA and VoiceOver typically skip it silently. JAWS may announce it as "zero width space" on verbose configurations.

Hangul Filler. Announced as "Hangul filler" or as a Korean phonetic placeholder, depending on the screen reader's Korean language pack.

Braille Pattern Blank. Often announced as "blank" or "Braille blank cell." The most likely character to produce an unexpected spoken word.

Combining Grapheme Joiner. Almost always skipped silently. The least disruptive choice for accessibility.

The pragmatic guidance: a single invisible character for a blank bio is fine for almost everyone. Filling fields with many invisible characters degrades the experience for screen-reader users and is treated as hostile design by most accessibility audits.

When and where they get stripped at the API or database level

The stripping pipeline runs in three places, and understanding which one removed your character determines what to do about it.

Client-side input filters. JavaScript that runs before submit. Often strips Zero-Width Space and CGJ. Fix: paste the character rather than type it, since paste can bypass the input event listener.

Server-side Unicode normalization. NFC, NFD, NFKC, or NFKD applied on save. NFKC and NFKD are aggressive and remove many invisible characters. NFC is gentler and usually keeps Hangul Filler. Fix: use Hangul Filler over ZWSP.

Database trim and validation. SQL trim functions strip whitespace including some Unicode space characters. Validation libraries that reject "empty after trim" remove invisible characters before checking. Fix: use Hangul Filler, which is not whitespace and survives trim.

Use the tool

The TextKit Invisible Text Generator handles all five characters in one place. Pick the character type, set how many to generate, copy, and paste. The verify toggle reveals the codepoints so you can confirm the field is not actually empty before you save. All processing is local. Nothing is uploaded.

For typical use cases. One Hangul Filler for an Instagram bio. Two or three for a Discord display name. One for a WhatsApp blank message. Past five or six characters in a single field, you are almost certainly degrading accessibility without adding to the cosmetic effect.

The workflow that works

1. Pick the platform first. Instagram bio, Discord status, WhatsApp message, gaming username. Each platform has a slightly different sanitizer.
2. Default to Hangul Filler. It survives the largest number of sanitizers. Start here unless you have a specific reason not to.
3. Generate one or two characters. Not ten. The cosmetic effect is the same and accessibility is preserved.
4. Paste, do not type. Client-side input filters often run on the type event. Paste sometimes bypasses them.
5. Verify. Refresh the page or check from another account. If the bio is empty after save, the platform stripped the character. Try a different character (ZWSP if Hangul Filler was the first try, or vice versa).
6. If everything fails, use Em Space. It is visible (a wide blank) but survives almost every filter. Acceptable when nothing else works.

For related topics: Character Count: The Complete Guide explains how Unicode characters count against limits across platforms. Text Tools for the AI Era covers the broader Unicode tooling category. Case Converter and Character Counter are the companion tools that handle the rest of a Unicode workflow.