HTML Entity Decoder

Decode HTML entities into plain text and inspect one encoding layer. Compare named and numeric references before using text extracted from a webpage.

Open the Converter

Your input stays in this browser. Nothing is uploaded or saved.

Text limit: 1 MiB; XML/YAML nesting limit: 80 levels.

By YiyaoPublished Updated

Decode HTML entities without rendering the page

Paste text containing HTML character references and inspect the decoded characters. Use this tool when a collected field displays & instead of an ampersand, or when you need to compare encoded text with its readable form.

The result remains plain text. A decoded tag is shown literally rather than inserted into a webpage. This lets you inspect the characters without confusing a conversion result with a rendered document.

Identify the representation first

HTML entities, percent-encoded URLs and JSON string escapes use different rules. Choose the converter that matches the actual source field. A value containing %20 belongs to URL decoding, while a quoted JSON string may need JSON unescape.

Do not run every decoder in sequence. Record what the input represents and which transformation the downstream task requires.

Encoded exampleDecoded resultMeaning
A & BA & BNamed character reference
AADecimal character reference
AAHexadecimal character reference
&lt;item&gt;<item>Text that looks like markup stays text

Decode one layer and review it

Choose Decode once, compare the source and result, and copy the text when the change is correct. The input &amp;lt; becomes &lt; after one pass. It does not immediately become a less-than sign.

Keep the original source beside the decoded field in a test record. If an extraction library already decodes entities, an additional pass could change content that was intentionally encoded.

Decode one entity layer while keeping the result as plain text.

Keep text extraction separate from entity decoding

Entity decoding changes character references. It does not remove tags, select an article body, extract a price or validate a web page. If you have a full HTML document, first choose a suitable extraction method and identify the field you need.

For example, a product-title field containing Research &amp; Development becomes Research & Development. That operation says nothing about whether the surrounding page was collected completely or whether the title belongs to the intended record.

Use the tool in a collection workflow

Review the actual output from Arcvenza Crawler Console before transforming it. General Website Crawl offers HTML or Markdown; identify which form you collected and what your downstream field mapping expects.

Apply decoding only to the necessary text fields, preserve the source URL and inspect a few records manually. Continue to a structured table after extraction and validation, not by treating raw HTML as a JSON record list.

For collection tasks, Explore Crawler Console. Local conversion remains a separate step.

Check the edge cases

Unknown references remain visible so they can be investigated. Invalid numeric references follow the documented HTML decoding rules and can produce replacement characters. Treat those results as a reason to inspect the original source, not as corrected business data.

Whitespace can also carry meaning. A non-breaking space and an ordinary space may look similar while behaving differently in comparisons. Keep normalization as a separate, explicit decision.

HTML decoding questions

Does this remove HTML tags? No. It decodes character references and displays the result as text.

Is decoding the same as sanitization? No. Applications still need safe handling for the final output context.

Can it load a private page for me? Paste a sanitized text sample. This version does not fetch remote URLs.