Markdown Word Counter

How to use

1. 1Paste or type Markdown into the input. Load sample fills the editor with a small post if you want to see the counts before pasting your own.
2. 2Pick a counting mode chip: Blog post for everyday content, Strict prose for narrative only, Count everything for parity with a plain word counter, or Text to speech for voice scripts.
3. 3Toggle the Strip rules to customize a mode. Frontmatter, fenced code, indented code, inline code, link URLs, image syntax, HTML tags, footnotes, tables, and alt text can each be flipped on or off.
4. 4Adjust reading and speaking words-per-minute if your audience reads or speaks faster or slower than the defaults.
5. 5Read the headline cards (prose words, source words, reading time, speaking time) and the document structure panel. The per-section table breaks the prose count down by heading.
6. 6Use Copy summary to copy the full numeric report, or Copy CSV to export the per-section breakdown into a spreadsheet.

About this tool

Markdown Word Counter measures Markdown the way readers experience it. A plain word counter treats every token in the source as a word, so URLs, fenced code blocks, frontmatter keys, image syntax, footnote definitions, emphasis markers, table separators, and HTML tags all inflate the count. Substack, Medium, Hugo, Jekyll, Ghost, MkDocs, Docusaurus, and most static-site generators compute reading time on the prose only, and that is the number writers care about. This tool reports both: the raw source word count for the unprocessed file, and the prose word count after stripping the Markdown syntax you select. Counting is governed by ten transparent toggles: strip frontmatter (YAML --- ... --- and TOML +++ ... +++), skip fenced code blocks (``` and ~~~), skip 4-space indented code blocks, skip inline code spans, drop link URLs while keeping link text, skip image syntax (with or without alt text), strip HTML tags, skip footnote definitions and markers, skip pipe tables, and count image alt text. Four presets cover the common workflows: Blog post (the default, drops code and URLs but keeps inline code and link text), Strict prose (drops everything that is not narrative text), Count everything (matches a plain word counter for parity), and Text to speech (a clean voice script with code, URLs, images, tables, and HTML removed). Beyond word counts the tool reports document structure stats: characters with and without spaces, sentences, paragraphs, source lines, heading counts per level H1 to H6, link count, image count, image alt word count, list items, GitHub task list items (with done count), blockquote lines, fenced code blocks (count, lines, characters), inline code spans, pipe tables, HTML tags, and footnotes. Reading time uses your configurable speed (default 225 words per minute for silent reading) and a separate speaking time uses a slower rate (default 150 wpm) for podcast and voiceover prep. A per-section breakdown lists every ATX or Setext heading with its prose word count, indented by level so the document outline is visible at a glance, and a Copy CSV button exports the table for editing in a spreadsheet. The full numeric summary can be copied to clipboard with one click. Everything runs in your browser; the Markdown is parsed locally so the source never leaves your device. Useful for blog authors hitting a word target, technical writers measuring README density, ghostwriters quoting per-word jobs, documentation teams enforcing length budgets, podcast hosts checking script length, content briefs comparing draft size, and any workflow where the source word count and the reader word count must match.

Free to use. Works in your browser. No signup, no login.