<> HTML to Markdown Converter, Clean Markdown From Any HTML

What is the HTML to Markdown Converter?

Markdown and HTML describe the same document in two different ways. HTML is verbose and explicit, wrapping every heading, paragraph, and link in angle-bracket tags. Markdown is a lightweight shorthand where a "#" means a heading and "*word*" means bold. Converting from HTML to Markdown means reading the structure the tags describe and writing the shortest text that reproduces that structure. This tool does exactly that: it does not use regular expressions to hack at your HTML string, it builds a real document object model and walks it node by node.

The parsing step matters more than it looks. When you call new DOMParser().parseFromString(input, "text/html"), the browser applies the same forgiving HTML rules a web page uses, so unclosed tags, stray whitespace, and nested inline elements are all normalized into a clean tree before any conversion happens. The converter then recurses through the body: block-level elements like h1 to h6, p, ul, ol, blockquote, pre, and table each become their own Markdown block separated by a blank line, while inline elements like strong, em, a, code, and img are rendered in place inside the text of their parent block.

Table conversion deserves special mention because it trips up a lot of simple converters. Standard Markdown, known as CommonMark, has no table syntax at all; the pipe-and-dash table format you see on GitHub is a GitHub Flavored Markdown, or GFM, extension. This tool reads each table row, uses the first row as the header, inserts the required "---" separator row, and pads each cell so the output is valid GFM. That makes it a genuine html table to markdown table converter, not just a heading and list converter, which matters if you are pulling pricing tables, comparison charts, or spec sheets out of a web page or a WordPress export.

A few rules keep the rest of the output tidy and portable. Runs of spaces, tabs, and newlines inside text are collapsed to single spaces, because in HTML that whitespace is not significant but in a text file it would create ragged output. Characters that carry meaning in Markdown, such as asterisks, backticks, and square brackets, are escaped in plain text so a literal asterisk in your content does not accidentally turn into italics. Inline code and fenced code blocks are the exception: their contents are copied verbatim so your code is never altered. Lists support simple nesting, indenting child items by two spaces per level, and ordered lists number themselves sequentially.

The result is Markdown that renders identically to your source in any CommonMark or GitHub Flavored Markdown viewer, minus the tags. It is ideal for pulling content out of a CMS or a rich text editor, cleaning up pasted HTML, or migrating a page into a Markdown-based system like a static site generator, an Obsidian vault, or a documentation tool. Because the whole conversion happens client side in plain JavaScript, there is no upload, no size limit imposed by a server, and no waiting on a round trip.

Worth knowing before you migrate a large batch of pages: Markdown flavors are not perfectly interchangeable. GitHub, Obsidian, and most static site generators accept GFM tables and fenced code without complaint, but Notion strips YAML frontmatter on import and renders some nested formatting differently, and plain CommonMark viewers will not render a pipe table at all. If your destination is one of those platforms, check its docs for what it actually supports before assuming the Markdown this tool produces will look identical everywhere.

When to use it

• Migrating articles or pages out of a CMS or WordPress export and into a Markdown-based static site like Astro, Hugo, or Jekyll.
• Cleaning up HTML copied from a rich text editor or email into portable Markdown for a README or docs page.
• Converting the HTML output of an API or scraper into Markdown for storage in Git or a knowledge base.
• Turning a styled web snippet into plain Markdown so it renders consistently on GitHub, GitLab, or a wiki.
• Converting an HTML pricing table, spec sheet, or comparison chart into a GitHub Flavored Markdown table.
• Preparing content for a Markdown-driven chat app, note app, or Obsidian vault by stripping HTML down to its structural essentials.

How to use the HTML to Markdown Converter

1. Paste or type your HTML into the input box at the top.
2. Choose whether unordered lists use "-" or "*" with the bullet-style checkbox.
3. Read the converted Markdown in the output box, which updates live as you edit the input.
4. Check any tables in the output render correctly as pipe tables if your source HTML included a table.
5. Click the Copy button to copy the Markdown to your clipboard.
6. Paste the Markdown into your README, static site, wiki, or docs tool.

Formula & method

Worked examples

How each HTML element maps to Markdown

Markdown characters escaped in plain text (but not inside code)

Where the output Markdown will and will not render as expected

Common mistakes to avoid

• Expecting inline CSS and classes to survive. Markdown has no concept of colors, fonts, or class names. Style attributes, class attributes, and span wrappers are dropped, and only the structural meaning of the HTML is kept. If a look depends on CSS, that look will not carry over to Markdown.
• Pasting a full page with head, scripts, and styles. The converter walks the document body, so navigation, script tags, and style blocks are ignored rather than turned into text. Paste the content region you actually want, not an entire saved web page, for the cleanest result.
• Assuming every Markdown flavor renders GFM tables. Pipe tables are a GitHub Flavored Markdown extension, not part of the core CommonMark spec. This tool outputs GFM tables because that is what GitHub, GitLab, most static site generators, and Obsidian expect, but a strict CommonMark-only renderer will show the pipes and dashes as literal text.
• Losing whitespace inside code you meant to keep. Only inline code and fenced pre blocks preserve exact spacing. Whitespace in ordinary paragraphs is collapsed to single spaces on purpose, so if indentation matters, wrap that content in a pre or code element before converting.
• Expecting merged or nested table cells to convert cleanly. GFM pipe tables have no concept of colspan, rowspan, or a cell containing its own nested table. Complex HTML tables with merged cells are flattened as best as possible, but visually complex tables usually need a manual pass after conversion.
• Importing the result into Notion and expecting frontmatter or exact formatting to survive. Notion strips YAML frontmatter on import with a warning, and it can turn an imported table into a database rather than an inline table. If your destination is Notion, paste the Markdown into an existing page instead of importing it as a file, and re-check tables afterward.

Glossary

Markdown

A lightweight plain-text formatting syntax where symbols like "#" and "**" stand in for HTML tags, designed to be readable as-is and to convert to HTML.

HTML

HyperText Markup Language, the tag-based language browsers use to describe the structure and content of a web page.

DOMParser

A built-in browser API that parses an HTML or XML string into a document object model tree that code can walk and inspect.

CommonMark

The standardized core Markdown spec that most parsers agree on, covering headings, emphasis, lists, links, and code, but not tables.

GFM (GitHub Flavored Markdown)

An extension of CommonMark used by GitHub that adds pipe tables, task lists, and strikethrough on top of the core spec.

Block element

An element such as a heading, paragraph, list, table, or blockquote that forms its own line or region, separated from other blocks by a blank line in Markdown.

Inline element

An element such as strong, em, a, or code that lives inside the text of a block and formats a span of characters.

Fenced code block

A block of code wrapped in triple backticks in Markdown, which preserves spacing and disables all other formatting inside it.

Frequently asked questions