JSON Console Logo
JSONconsole

"category": "dev-utility",

HTML Unescape Tool

"tldr": Paste text full of &, <, — and read it as plain text - named, decimal, and hex entities all decoded.

Decode HTML entities back into readable characters: named entities (&, ©, —), decimal references (—), and hex references (🎉) all become their real characters, emoji included.

{"html entities": "text"}

Entity-encoded text shows up wherever HTML was flattened into data: scraped pages, RSS feeds, CMS exports, and APIs that double-encode. When your UI displays "Fish & Chips", this is the tool that shows you how many layers deep the encoding goes.

How to decode HTML entities

  1. 1Paste the entity-encoded text.
  2. 2Named (&), decimal (—), and hex (🎉) entities are decoded; unknown sequences pass through untouched.
  3. 3If the output still contains entities, the source was double-encoded - run it again.
  4. 4Copy the clean text.

Convert HTML entities to Text in code

Python
import html

html.unescape("Fish & Chips — ©2026")
# 'Fish & Chips — ©2026'
JavaScript (browser)
const unescapeHtml = (s) => {
  const el = document.createElement("textarea");
  el.innerHTML = s;
  return el.value;
};

Frequently asked questions

Why does my text say & - what happened?

Double encoding: "&" was escaped to "&", then that string was escaped again, turning its & into & - yielding &. Each pipeline stage escaped without checking. Decode twice here to read it; fix it upstream by escaping exactly once, at final output.

Will decoding scraped content make it safe to display?

The opposite - decoding turns inert &lt;script&gt; text back into live <script> markup. Decode to *read* or process text, but if you re-display user-derived content in HTML, escape it again at output. Decoding is for humans and data pipelines, not a sanitizer.

Which named entities are supported?

The ones that appear in real-world data: the XML five plus the common typography and symbol set (dashes, quotes, ©®™, currency, «», °±×÷µ…). All numeric references work universally, so anything exotic still decodes via &#NNNN; form.

Last updated:

You might also need

Free and in-browser, like everything on JSON Console. Browse all tools →

HTML EscapePaste text or code and get it HTML-escapedURL DecodePaste a percent-encoded URL or parameter and read it as plain textJSON UnescapePaste an escaped JSON string and get readable textURL EncodePaste text and get it percent-encoded for safe use in URLsBase64 EncodePaste any text and get Base64Base64 DecodePaste Base64 and read it as textMarkdown to HTMLPaste Markdown and get clean HTMLCase ConverterPaste identifiers (one per line) and get every casing at once: camelCase, PascalCase, snake_case, kebab-cas...UUID GeneratorGenerate up to 1,000 UUIDs at once