HTML Entity Encoder & Decoder

Last Updated: August 2026

🔴 What Each Tab Actually Does

This page started life as a plain html entity encoder with four output formats. It now carries four tabs, because in practice the encoding job and the “why is my pasted code broken” job turn up in the same ten minutes of work. Here is what each tab is for and when to reach for it.

Convert — nine escape formats

Paste text on the left, pick a format, read the result on the right. The conversion runs on every keystroke, so there is no button to press. Switch the Encode / Decode toggle to run it backwards.

A real example. Drop this line into the input with HTML Entities — essential characters selected:

<p class="note">Price: 12 & rising</p>

and you get back:

<p class="note">Price: 12 & rising</p>

Drop that output into a page and the reader sees the tag, rather than the browser rendering an empty paragraph. That is the whole point.

Inspector — one row per character

Paste anything and every character gets a row: its Unicode name, its codepoint, its decimal value, its UTF-8 byte sequence, and its HTML, CSS and JavaScript escapes side by side. Type a single emoji and you get one row, not two broken halves — the table walks codepoints rather than JavaScript string units, so 😀 is reported correctly as U+1F600 with four UTF-8 bytes.

The byte column is the one people find useful unexpectedly. A database column set to 20 bytes holds 20 plain Latin letters, but only 6 emoji. If a field keeps truncating, check the byte count here before blaming the database.

Cleaner — the invisible character scanner

This is the tab worth bookmarking. Paste text copied from Word, Google Docs, a PDF or a chat window, and the scanner counts what is hiding inside it: zero-width spaces, byte order marks, non-breaking spaces, curly quotes, en and em dashes, and bidirectional override marks. The preview marks every one of them so you can see exactly where they sit.

Then tick the fixes you want and press Clean Text. Each fix is separate, because sometimes you want the curly quotes gone but the em dashes kept — an editor cleaning prose has the opposite preference to a developer cleaning a code snippet.

Cheat Sheet — 130+ entities, searchable

Filter by category or search by name, description or number. Typing 8364 finds the euro sign, arrow returns the arrow block, and alpha jumps to Greek. Click any code to copy it.

🟡 Three Jobs This Page Handles Well

🔵 Publishing a code snippet in a blog post or CMS. Run it through Convert with HTML essential characters. Without that step, a WordPress editor swallows your markup and the reader sees a rendered box instead of the code you meant to show.

🟠 Fixing a script that fails on a line that looks perfect. Paste the failing line into Cleaner. Nine times out of ten it is a non-breaking space where a normal space should be, or a curly apostrophe that arrived with the copy. Both are invisible in most editors and both stop the parser dead.

🟣 Working out why a character renders as a box. Put it in Inspector, read the codepoint, and you know whether the data is wrong or the font simply has no glyph for it. Those two problems have completely different fixes, and the codepoint tells you which one you have.

🟢 Mistakes and Limits Worth Knowing

Escaping is not the same as sanitising

Encoding output before it reaches a page is a sound habit. Decoding untrusted input is not a security control, and this tool is not a filter for hostile markup. If you are handling user-submitted content on a live site, escape at the point of output in your own code — do not paste user input through a browser tool and treat the result as safe.

Each format belongs to one context

Named entities like   are defined by HTML, not by XML. Paste them into an RSS feed, an SVG file or a raw XML document and the parser throws an undefined-entity error. XML defines exactly five: <, >, &, " and '. That is why the XML option in the format list exists as its own entry, and it is the one to pick whenever the destination is not an HTML page.

What this page deliberately does not do

🔵 Percent-encoding. If you need %3C and %20 for a URL or query string, that lives in the Universal URL Encoder & Decoder. Percent-encoding and HTML entities solve different problems and mixing them produces double-encoded strings that display as literal < on the page.

🟠 Base64. That belongs in the Base64 Encoder / Decoder. Base64 makes binary data safe to transport as text; entities make reserved characters safe to display. Different jobs.

🟣 Bulk find and replace across many files. The Cleaner works on whatever you paste into it. For a pattern-driven sweep across a large body of text, the Regex Tester & Debugger is the better fit.

Performance ceilings, honestly stated

The Inspector renders the first 400 characters and tells you when it has truncated, because a table with fifty thousand rows locks up the tab on a mid-range phone. The Cleaner highlights the first 3,000 characters in its preview, though the count and the actual cleaning run over the whole input. Convert has no practical limit short of the browser’s own memory, but pasting a multi-megabyte file into any browser textarea will feel slow no matter which site you are on.

🔴 Where to Go Next

If you want the reasoning underneath all of this — how Unicode codepoints differ from UTF-8 bytes, why the HTML parser needs entities at all, and how invisible characters became a genuine software supply-chain problem — that is covered in the companion piece: HTML Entity Encoder & Decoder: the complete guide.

The formal definition of the named character references lives in the WHATWG HTML specification, and MDN’s character reference glossary is the friendlier read of the two.

❓ Frequently Asked Questions