Markdown Extract Bold Text

Collect every bold phrase from a Markdown document

Writers bold the words that matter: key terms, warnings, product names, decisions, deadlines. In a long Markdown file those highlights are scattered across dozens of paragraphs. The Markdown Extract Bold Text tool gathers them into one list. Paste your Markdown, press Extract, and you get a JSON array of every phrase wrapped in double asterisks or double underscores, in the order they appear.

The result is a fast way to see what a document emphasises without rereading it, and a handy starting point for glossaries, summaries, flashcards and style checks.

What the tool recognises

Two bold syntaxes are supported, matching the "Supported formats" note on the page:

• **double asterisks**
• __double underscores__

The text between the markers must not start or end with a space. ** loose ** is not counted, which mirrors how Markdown renderers treat it. Leading and trailing whitespace is trimmed from each match, and empty matches are skipped.

Two places are skipped on purpose, so code samples do not pollute the list:

• Fenced code blocks opened with ``` or ~~~. Everything until the closing fence is ignored.
• Inline code between single backticks, such as `"**not extracted**"`.

The built-in sample, loaded with Load sample, demonstrates all of this: it contains bold text in normal paragraphs, bold markers inside inline code, and bold markers inside a JavaScript code block. Only the three genuine bold phrases come out.

A worked example

Given this Markdown:

The **deadline** is Friday. Use __staging__ first.
Run `echo "**hi**"` to test.

The output is:

["deadline", "staging"]

Duplicates are kept. If "deadline" is bold three times, it appears three times, which tells you how often a term is emphasised. Remove duplicates in your own script or spreadsheet if you need a unique list.

Step-by-step instructions

1. Paste your document into the Input Markdown editor, or press Paste to read from the clipboard. Not sure what to try? Press Load sample.
2. Watch the three counters at the top update as you type: Input lines, Found items and Code blocks skipped. Found items is a live preview of how many bold phrases will be extracted.
3. Press Extract, either in the header or under the editor.
4. Review the JSON array in Extracted Output.
5. Use Copy to put the array on your clipboard, Download to save it as markdown-bold-text.json, or Select to highlight the whole output for manual copying.
6. Press Clear to empty both the input and the output.

Reading the three counters

The counters at the top of the page update on every keystroke, before you press Extract, so they double as a quick diagnostic.

• Input lines counts every line in the editor, including blank ones. It confirms that a long paste arrived in full.
• Found items runs the same scan the Extract button uses and shows how many bold phrases it would return. If it reads zero on a document you know has bold text, the markers are probably padded with spaces or split across lines.
• Code blocks skipped counts the fenced blocks that were ignored. If this number is higher than you expect, an unclosed fence may be hiding the rest of the document from the scan, because everything after an opening fence is treated as code until a closing fence appears.

Private by design

Scanning happens in your browser with JavaScript. The Markdown you paste is not sent to this site's server, so you can safely run internal notes, meeting minutes or unpublished drafts through it. There is no server-side size limit.

Ideas for using the output

Build a glossary or index

Many style guides bold a term the first time it is defined. Extracting bold text from a manual gives you a ready-made list of defined terms to turn into a glossary.

Make study cards

Lecture notes written in Markdown often bold the key concepts. The array can be imported into a flashcard app or a spreadsheet as the front of each card.

Pull action items from meeting notes

If your team bolds owners, decisions or dates, one extraction produces the list of things that need follow-up.

Check emphasis in a draft

Editors often warn against bolding too much. A long list of bold phrases in a short article is a sign that nothing stands out any more. Seeing the list in one place makes that obvious.

Feed other tools

Because the output is valid JSON, it plugs straight into scripts, keyword research sheets or prompts for a language model that should focus on the highlighted points.

Tips for better results

• Keep bold phrases on a single line. The scan works line by line, so a phrase whose closing markers sit on the next line is not matched.
• Use one bold style consistently in your own writing. Both are supported, but consistency makes the source easier to maintain.
• If the Found items counter shows zero while you can see bold text, check for spaces just inside the markers.
• Compare Found items with the length of the JSON array after extraction. They use the same scan, so they should match.

Turning the array into something else

The JSON array is deliberately simple: a list of strings in quotes, separated by commas. Most spreadsheet apps can split it into rows after you remove the square brackets, and any programming language can read it with its standard JSON parser. If you just need a plain list to paste into a document, copy the output and delete the brackets, quotes and commas with a find-and-replace.

What not to expect

• Single-marker emphasis, *italic* or _italic_, is not extracted. This tool is only about bold.
• HTML bold tags such as <strong> or <b> are not detected.
• Inner Markdown is not cleaned. A bold link like **[Docs](https://example.com)** comes out as the raw text [Docs](https://example.com), and bold italic written with three asterisks keeps one asterisk on each side.
• Bold text inside a fenced code block or inline code is always skipped, even if you wanted it.

Limitations

• Matching is per line; multi-line bold spans are missed.
• No de-duplication, sorting or counting per phrase; the array lists matches in document order.
• Only JSON output is offered. There is no CSV or plain-list export, though the JSON is simple to convert.
• Indented code blocks without fences are scanned like normal text, so bold markers inside them can be picked up.

Related tools

Headings are the other main signal of structure in a document; the Markdown Heading Extractor lists them with levels and anchor slugs. To pull out the code samples this tool skips, use the Markdown Code Block Extractor.

Frequently Asked Questions