0 words·0 lines·0 chars·0 B·0 min read
Markdown
Previewrendered here, never sent anywhere

Free Online MD File Reader — Open and Preview Markdown in Your Browser

This page reads a .md file and shows it as formatted text: headings, paragraphs, bullet and numbered lists, GitHub-style task lists, tables with column alignment, blockquotes, horizontal rules, fenced code blocks, inline code, links and images. Click Open .md, drag a file anywhere onto the page, or paste Markdown straight into the left panel — the preview appears immediately, with no page reload and no server involved.

It is a reader, not an editor: nothing writes back to your file. Use it to skim a README.md, CHANGELOG.md, LICENSE-adjacent notes, exported Obsidian or Notion pages, or documentation shipped inside a downloaded archive — on any machine, including ones where you cannot install a Markdown viewer.

How To Read a Markdown File Online

Open a file: press Open .md (or Upload in the Markdown panel) and pick a .md, .markdown, .mdown, .mkd or .txt file. Or drag and drop it onto the page — a drop zone outline appears while you drag. Or paste text into the Markdown panel; the preview updates as you type, and Ctrl+Enter re-renders on demand.

Three views are available in the bar above the panels: Preview for the rendered document, Both for source and preview side by side, and Markdown for the raw text. In Both the divider between the panels is draggable — release it wherever you want the split, double-click it for an even 50/50, or select it with Tab and nudge it with the arrow keys; the proportion you settle on is remembered in this browser, and on a phone-width screen the panels stack instead. The bar also reports words, lines, characters, byte size and estimated reading time. When a document has headings, Contents opens an outline; clicking an entry scrolls the preview to that section, and in-document #anchor links work the same way.

Your document is kept in this browser, so a refresh does not lose it. Clear All empties the panels and deletes that saved copy; Copy Markdown puts the source back on your clipboard; Print (or Ctrl+P) puts the rendered document on paper or in a PDF file. Printing leaves out the panels, the view buttons, the navigation and this page's own text, prints dark-on-white whatever theme you read in, and keeps tables and code blocks whole across page breaks.

Which Markdown Syntax Is Supported

The renderer covers the CommonMark blocks people actually use plus the GitHub Flavored Markdown extensions that appear in repositories:

  • Headings — ATX (# … ######) and setext (underlined with === or ---), each given an id for the outline.
  • Lists — nested bullet and numbered lists, correct indentation handling, 1. start numbers, and - [x] / - [ ] task checkboxes.
  • Tables — pipe tables with :---, ---: and :---: alignment, escaped \| inside a cell, and inline markup in cells.
  • Code — fenced with ``` or ~~~ plus a language label, and four-space indented blocks; blank lines inside a block are preserved.
  • Inline — *, **, _, __, ~~ emphasis, inline code spans, backslash escapes, two-space hard line breaks.
  • Links — inline, reference and shortcut forms, autolinks, e-mail autolinks and bare URLs; images with alt text and titles.

Not interpreted: raw HTML (see below), YAML front matter, which is displayed as text because it is metadata rather than document content, and footnotes. HTML comments are removed.

Why Raw HTML Inside a Markdown File Is Shown, Not Rendered

Many Markdown "viewers" pass HTML through untouched, which turns a README file into an execution surface: a hostile repository can ship a badge image or an inline <script> that runs the moment you preview it, and link destinations written as javascript: or data: URLs do the same when clicked.

This reader takes the opposite stance. The parser escapes text character by character while it tokenises, so <b>bold</b> inside your file is displayed exactly as written and can never close a tag or open one this page would honour. Link and image destinations are checked against a scheme allowlist — http, https, mailto and tel — and anything else keeps its visible text but loses the link. External links are opened with rel="noopener noreferrer".

Images, Relative Paths and Privacy

An image written with an absolute URL loads normally. An image written with a relative path — ![diagram](./docs/arch.png) — cannot: your file's folder is never uploaded, so there is nothing next to this page for that path to resolve against. The alt text and title are shown instead, and the source panel always keeps the original line. That is the deliberate trade-off for a reader that never transmits your files.

Everything runs inside your browser tab: the Markdown parser, the outline builder and the file reader are plain JavaScript modules delivered with this page. There is no upload endpoint, no conversion API and no analytics on your document content. Files up to 5 MB are accepted; larger ones are rejected with a clear message rather than freezing the tab.

Frequently Asked Questions

Everything you need to know about reading Markdown files in a browser.

How do I open an .md file in my browser?

Click Open .md in the top bar and choose the file, drag it onto the page (a dashed drop zone appears while you drag), or paste Markdown into the Markdown panel. The preview appears immediately — no reload, and the extension is checked, so .md, .markdown, .mdown, .mkd and .txt all work.

The Preview / Both / Markdown switch decides whether you see only the rendered document, both panels side by side, or only the source. In Both the preview follows every keystroke, which makes it useful for checking a table or a code fence while you fix it.

In Both you also choose the split: drag the divider between the panels to wherever is comfortable, double-click it for an even 50/50, or select it with Tab and use the arrow keys. The width you settle on is remembered in this browser. Below roughly 640 pixels the panels stack, because two side-by-side columns that narrow stop being readable.

Can I print the Markdown file after opening it?

Yes — press Print in the top bar, or just Ctrl+P (Cmd+P). Both open your browser's normal print dialog, so you can send the document to a printer or choose Save as PDF.

What reaches the paper is the rendered document only: the Markdown panel, the view buttons, the stats bar, the outline, the site navigation and this page's own explanatory text are all left out, and the 70%-height scroll box opens up so the text flows across sheets instead of being cut at the fold.

The sheet is printed dark ink on white whatever theme you read in, long code lines wrap instead of running off the edge, and tables, code blocks and blockquotes are kept whole rather than split mid-row. Nothing is uploaded: printing is done by your browser, and an empty panel simply refuses with a note in the status bar.

Does this MD reader upload my file to a server?

No — nothing leaves your machine. The file is read by your own browser through the File API and parsed by a plain JavaScript module that was already delivered with this page (/js/md.96db30916b.mjs). There is no upload endpoint, no server-side converter, and no analytics that inspects your content.

That is also why relative image paths cannot be resolved: the reader only ever has the single file you picked, never the folder around it. If a viewer promises to render a whole repository tree, it is by definition fetching it somewhere else.

Which Markdown syntax does it support?

The CommonMark block set plus the GitHub Flavored Markdown extensions that appear in real repositories: ATX and setext headings, paragraphs with two-space hard breaks, nested bullet and numbered lists, - [x] task checkboxes, pipe tables with alignment markers, fenced (``` / ~~~) and four-space indented code blocks with a language chip, blockquotes, horizontal rules, emphasis and strikethrough, inline code spans, backslash escapes, images, reference and shortcut links, autolinks and bare URLs.

Not interpreted: raw HTML, YAML front matter (shown as text, since it is metadata), and footnotes. HTML comments are dropped. Each heading gets an id prefixed with md-, which is what lets the outline and in-document #anchor links scroll the preview without colliding with the page's own element ids.

Why isn't the HTML inside my Markdown file rendered?

Because a Markdown viewer is a common way to get code executed by accident. A README can embed <img onerror=…>, an inline <script>, or a link whose destination is javascript: — and a viewer that passes HTML through runs it as soon as you preview the file.

This parser escapes text character by character while tokenising, so tags stay literal: <b>bold</b> appears exactly as written. Destinations go through a scheme allowlist (http, https, mailto, tel); a refused destination keeps its text and drops the link. External links carry rel="noopener noreferrer".

Why are the images in my Markdown file not showing?

Images addressed by absolute URL load normally. Images addressed by relative path — ![diagram](./docs/arch.png), or an attachment path exported from Obsidian or Notion — cannot: only the .md file is read, so there is no folder next to this page for that path to resolve against.

The alt text and title are shown in place of the picture, and the original line stays in the Markdown panel, so nothing in your document is lost. To carry the images along, host them or export them with absolute URLs.

Can I read a README straight from GitHub or another website?

By content, not by URL: open the Raw view on the host, copy everything, and paste it into the Markdown panel — or download the file and use Open .md. The preview handles READMEs with badges, tables, task lists and code fences.

Fetching a URL here would need a proxy on a server and would disclose the address you wanted to read, so this page deliberately has no "load from URL" field.

Is my Markdown document saved anywhere?

Yes — in your own browser, in localStorage, so closing the tab or reloading does not lose a long document you pasted. The file name is kept beside it only to label the panel.

Nothing is transmitted, and Clear All removes both stored entries along with the panels. On a shared or borrowed computer, press it before you leave; on your own machine the saved copy is usually handy the next time you need to re-read something.