HTML tag stripper

This tool is written for raw HTML source: view-source output, scraped pages and feed exports, rather than text that merely picked up incidental formatting on the way into a clipboard. It removes every tag and the attributes inside it, drops script and style blocks with their contents, and decodes the entities that account for most of what actually appears in ordinary markup.

Processing happens client-side, in JavaScript, in your browser. There is no request sent anywhere to log or queue, which is also why there is no run limit: nothing here has a server cost to meter.

No configuration, and no step gated on network latency.

Step 1

Paste the raw source into the input field.

Step 2

Run the strip. It executes instantly, locally.

Step 3

Read the output and check for any unusual leftover entity.

Typical markup carries three distinct categories that this tool handles individually, even though the visible result is the same removal for all three. Every tag, opening and closing, is stripped. Every attribute living inside a tag, class, id, inline style, data attributes, hrefs, is removed with the tag itself, since none of it was ever rendered as visible text.

Script and style blocks are removed wholesale, tags and contents together, because JavaScript and CSS were never reader-facing content. A page that is mostly boilerplate and tracking scripts can reduce to a small fraction of its source length once stripped, which is frequently the reason someone runs this in the first place.

An entity is a character represented in code because the literal character would be ambiguous inside markup: & for an ampersand, < and > for the brackets a literal less-than or greater-than sign would collide with.

Four are decoded automatically:   to a space, & to &, and < and > to their literal characters. That set covers the overwhelming majority of what shows up in real-world source. Numeric entities such as ’ and named entities such as © or — are left exactly as written, since there is no bounded list of every entity a given source might contain. Inspect the output for these if the source used typographic punctuation.

It matches and removes tag patterns; it does not construct a document object model the way a browser engine does. On well-formed markup, the outcome matches what a full parser would produce. On malformed markup, or on plain text containing a stray angle bracket used as a mathematical comparison, a regex-based pass can diverge from a parser's result. Review output on anything structurally unusual before relying on it downstream.

Run raw source through a word counter, a translation step or a similarity check and every tag name and attribute value is counted as content, skewing the result. In structured data it is a harder failure: a tag sitting unescaped inside a scraped CSV field or a JSON string can break parsing entirely, and the resulting error rarely identifies a leftover tag as the cause.

Engineers cleaning scraped pages ahead of a pipeline, a search index or a model use this as a standard preprocessing step. Teams working with RSS and Atom feeds, or migrating legacy CMS exports, use it to extract readable text without carrying markup forward.

It is also used in QA, checking a template's rendered output for a stray unescaped tag or entity before it reaches real readers.