Why identical-looking text can differ
A non-breaking space can look like an ordinary space. A zero-width space can sit between letters without a visible gap. Exact matching still treats them as different values.
Find first, then change
Paste the text, inspect the character codes and positions, and choose an action for each unwanted type. Preview the result before copying or downloading. The original input stays intact.
Some hidden characters belong there
Join controls and variation selectors can matter to writing systems and emoji. The finder identifies them but preserves them. Directional controls may be valid in multilingual text; removing the whole group changes display behavior.
This is a check against a defined Unicode set, not a guarantee that text is harmless or compatible with every application. It does not decode literal escape sequences, normalize accents or strip HTML.
Data: Unicode 17.0.0. References: Unicode character properties, emoji sequences, bidirectional text. Unicode data license.
Invisible character FAQ
What does this invisible character finder detect?
It checks a fixed Unicode 17.0.0 set of whitespace, default-ignorable code points, format and control characters, plus Braille blank. It also reports replacement characters and unpaired UTF-16 surrogates. Not every finding is an error or visually invisible in every context.
Why are ZWJ, ZWNJ and variation selectors preserved?
They can affect writing systems or combine and select the appearance of emoji. This version identifies them but does not offer removal. A default-ignorable property is not permission to delete a character.
Can I replace non-breaking spaces with normal spaces?
Yes. Choose Replace with ordinary space for the specific type and preview the result. This changes spacing and line-breaking behavior. Nothing is replaced by default.
Is my text uploaded or saved?
No. This tool processes your text in this browser, without uploading it or saving its contents. Copy and download happen only when you choose them.
Does cleaning guarantee that my text will import correctly?
No. Selected changes can help with exact matching and unwanted formatting, but your destination may have other requirements. Review the changed text before using it.
Can pasted text reveal its original encoding or BOM?
No. A pasted U+FEFF can be identified, but it does not establish the original file encoding or BOM. Browsers can normalize line endings when text is pasted into a text area.
What is preserved in the downloaded TXT?
The TXT contains the previewed string encoded as UTF-8, with no extra BOM or final newline added. Any initial U+FEFF you chose to keep is still encoded, and some applications may interpret those bytes as a BOM. Unselected values are not trimmed, normalized or converted.