To turn fancy text back into normal letters, either retype it as plain A–Z or paste it into a Unicode NFKC normalizer. The reason "paste as plain text" fails is that styled text like 𝗯𝗼𝗹𝗱 or 𝓼𝓬𝓻𝓲𝓹𝓽 isn't formatting applied to normal letters — each character is a different Unicode code point, so there's no formatting layer to strip away.
Why won't "paste as plain text" fix it?
"Paste as plain text" (or "Paste and Match Style") removes formatting — bold set with a B button, a font-family, a color, a hyperlink. That formatting is metadata stored alongside ordinary letters. Fancy Unicode text has none of that. When you copy 𝗛𝗲𝗹𝗹𝗼 from a generator, you aren't copying the letters H-e-l-l-o with a "bold" attribute attached. You're copying five distinct characters from Unicode's Mathematical Alphanumeric Symbols block — 𝗛 is code point U+1D5DB, a completely separate character from the normal H at U+0048.
So there is nothing for "plain text" to strip. The plainest possible version of 𝗛𝗲𝗹𝗹𝗼 is still 𝗛𝗲𝗹𝗹𝗼, because that's what the characters are. This is the same reason the text survives a copy-paste into any app in the first place — and the same reason it can show as boxes on some devices (why fancy text shows as boxes). It's real Unicode, not a font file.
What actually converts it back to normal?
You have two reliable options and one bulk option:
| Method | How it works | Best for |
|---|
| Retype it | Read the styled text, type the same letters normally | A word or short phrase |
| NFKC normalizer | Software folds each styled glyph to its plain "compatibility" equivalent | A full bio, caption, or paragraph |
| Strip in a spreadsheet/script | unicodedata.normalize("NFKC", text) in Python, or =UNICHAR/formula tools | Bulk cleanup of many rows |
The technical name for the automatic method is Unicode Normalization Form KC (NFKC). Every styled letter in the math-alphanumeric range carries a "compatibility decomposition" that points back to a plain letter, so an NFKC pass rewrites 𝗯𝗼𝗹𝗱 → bold, 𝘪𝘵𝘢𝘭𝘪𝘤 → italic, and 𝓼𝓬𝓻𝓲𝓹𝓽 → script automatically. Fullwidth text (Ａ Ｂ Ｃ) and Fraktur (𝔄 𝔅 ℭ) fold too. In Python that's a single line; many "remove fancy text" and "normalize Unicode" web tools are just an NFKC call behind a text box.
Will NFKC clean up every fancy style?
Mostly, but not everything. NFKC folds the styles people generate most often — bold, italic, bold-italic, script, Fraktur, monospace, sans-serif, fullwidth, and circled letters (Ⓐ → A) all map cleanly back to plain A–Z. A few decorative styles are deliberately not compatibility-equivalent to a plain letter, so a normalizer leaves them as-is:
- Negative-squared / enclosed-alphanumeric-supplement letters (🅰 🅱 🅾) — these live in the Enclosed Alphanumeric Supplement block and have no plain-letter decomposition, so they can be left untouched.
- Small-caps letters (ᴀ ʙ ᴄ) — most are borrowed from the phonetic (IPA) block and don't fold to capitals.
- Regional-indicator and symbol substitutions — anything that's really an emoji or symbol standing in for a letter.
- Zalgo / glitch text — the base letters are normal, but the stacked combining marks survive NFKC (they get reordered, not removed).
For those, retyping is the sure fix. If you only need the readable letters back and a stray decorative glyph remains, delete it by hand — it's usually one or two characters. Note the flip side: because normalizers exist, fancy text is easy for platforms to strip on their end too, which is part of why some feeds flatten it (does fancy text hurt Instagram reach).
The quickest fix for one line of text
If it's a display name, a single bio line, or a heading, don't reach for a tool at all — just read the styled version and retype the same words in your normal keyboard. It takes seconds and guarantees clean, screen-reader-friendly output. This matters because fancy Unicode is read poorly (or skipped) by assistive tech (screen readers and fancy text), so converting back to plain letters is often the accessible choice, not just a cosmetic one (are Unicode fonts accessible).
For anything longer — a whole caption you pasted from somewhere, a list of names, a spreadsheet column — use an NFKC normalizer so you're not retyping paragraphs. Paste in, copy the normalized result out, then eyeball it for any of the non-folding styles above.
Where does BoldlyType fit in?
BoldlyType is a formatter — it goes the other direction, turning your plain text into copy-paste Unicode styles for social bios, captions, and display names. It doesn't ship a "back to normal" button, because the honest fix is either retyping or a one-line NFKC pass, and we'd rather point you to how it actually works than pretend there's a proprietary trick. If you're formatting text in the first place, the text generator, bold text generator, and Instagram text formatter show each style side by side so you can see exactly which Unicode block a given look comes from — which is the same knowledge that tells you how to undo it. When you want it back to normal, retype it or run it through NFKC, and it'll fold to plain A–Z again.
