Developer tool

Markdown to Confluence Converter

Convert Markdown to Confluence in the format your site accepts. The Cloud editor no longer takes wiki markup, so the Cloud target copies rich text that pastes in as native headings, tables, code snippets and action items. For Data Center or Server, switch to escaped wiki markup.

Parsed in your browser; nothing you paste is sent to Kalcify.

Markdown

CommonMark plus GitHub tables, task lists, strikethrough, autolinks and footnotes. The output updates as you type.

Paste target

Confluence CloudRich-text pasteData Center / ServerWiki markup

Paste Markdown on the left, or load the example, to see the rich-text preview here.

Confluence Cloud

Paste Markdown into Confluence Cloud

Keep the target on Confluence Cloud and select Copy for Confluence Cloud.
Open the page, press E (or select Edit) and click where the content belongs.
Paste with Ctrl+V (⌘V). Avoid “paste as plain text” (Ctrl+Shift+V): the plain-text half of the clipboard is your original Markdown.

The copy uses only elements the Cloud editor has its own version of — the set Atlassian lists as its Markdown shortcuts: headings 1–6, bold, italic, strikethrough, inline code, bulleted and numbered lists, quotes, code snippets, dividers and links. Tables are written with header cells in the first row. A list made entirely of tasks is written in the markup Atlassian's editor uses for action items, and each code block carries its fence language in the data-language attribute that editor reads. Atlassian also auto-detects CSS, Go, HTML, Java, JavaScript, JSON, Python, Ruby, Rust, Shell, SQL and TypeScript in pasted code; for any other language pick it from the snippet's language menu.

Why pasting raw Markdown turns into a code block

Copy Markdown out of VS Code and the clipboard carries VS Code's own HTML: a block of monospaced text. Atlassian's editor schema has a rule for exactly that paste — a block styled white-space: pre or in a monospace font becomes a code snippet — so your ## Overview arrives as code, not as a heading. The Insert > Markup dialog that used to convert Markdown belonged to the legacy editor, which Atlassian fully retired on April 1, 2026; its documentation now says the cloud editor doesn't support markup. Pasting rendered rich text instead is the fix reported in Atlassian's Community for this exact problem, and it is what the Cloud target puts on your clipboard without a Markdown preview app open.

Data Center and Server

Markdown to Confluence wiki markup

Switch the target to Data Center / Server and select Copy wiki markup.
In the Confluence editor choose Insert > Markup and select Confluence wiki as the format.
Paste, check the preview in the dialog, and select Insert.

Data Center's dialog can also take Markdown directly. Wiki markup is the format for the places that accept nothing else — the Advanced tab of the link browser and custom space header, footer or sidebar content — and for the Cloud REST API below. Confluence converts inserted markup to rich text immediately; there is no wiki source left to edit afterwards.

How escaping works

{ } [ ] | are always escaped with a backslash: they open macros and links and separate table cells.
* _ + ^ ~ - are escaped unless they sit between two letters or digits, where Atlassian's renderer ignores them — snake_case stays readable. Inside inline code ({{…}}) they are always escaped, so code such as a*b*c stays literal.
??, a ! followed by text, a leading # and emoticon codes such as (y) and :) are escaped so they stay literal.
Formatting that touches a letter, such as **bold**text, is written in the braced form {*}bold{*}text that Atlassian documents for formatting part of a word.

REST API v2

Send the wiki markup through the Confluence REST API

Confluence Cloud's v2 REST API accepts wiki markup as a page body: when you create or update a page, blog post or comment, set body.representation to wiki and put the converted markup in body.value. That makes the wiki output the input for scripts too, which is why this tool has no separate storage-format (XHTML) output.

POST /wiki/api/v2/pages
{
  "spaceId": "123456",
  "title": "Release 2.4 checklist",
  "body": {
    "representation": "wiki",
    "value": "h2. Release 2.4 checklist\n\n..."
  }
}

The markup has to be a JSON string, so newlines become \n and every backslash doubles. Run the finished request body through the JSON formatter and validator before sending it.

Worked examples

Markdown in, Confluence out

A release checklist

Markdown

## Release 2.4 checklist

Deploy **after** the `db:migrate` step. Full notes: [changelog](https://example.com/changelog).

| Service | Owner | Status |
| --- | --- | --- |
| API | Dana | Ready |
| Worker | Lee | Waiting on {queue} |

```bash
./deploy.sh --env production
```

- [x] Freeze the release branch
- [ ] Announce in #releases

Wiki markup output

h2. Release 2.4 checklist

Deploy *after* the {{db:migrate}} step. Full notes: [changelog|https://example.com/changelog].

||Service||Owner||Status||
|API|Dana|Ready|
|Worker|Lee|Waiting on \{queue\}|

{code:language=bash}
./deploy.sh --env production
{code}

* \[x\] Freeze the release branch
* \[ \] Announce in #releases

Copied for Confluence Cloud, the same input carries 1 heading, 1 table, 1 code block, 2 tasks, 1 link. The table's first row is written as header cells, the code block keeps bash as its language, and because every item in the checklist is a task, the list is written as action items rather than bullets. In the wiki markup the literal {queue} becomes \{queue\} so Confluence does not read it as a macro, while --env inside the code block stays as typed: code macro bodies are not parsed. Wiki markup has no task list, so the two checklist items keep a [x] / [ ] prefix.

Boundary case: characters that are markup in wiki syntax

Markdown

Set {timeout} in [config], pass --force, keep snake_case, and delete `C:\Temp\*.log`.

Wiki markup output

Set \{timeout\} in \[config\], pass \-\-force, keep snake_case, and delete {{C:\Temp\*.log}}.

Braces, brackets and pipes are escaped wherever they appear. The double dash is escaped because wiki markup turns -- into an en dash, while the underscore in snake_case is left alone: a delimiter between two letters cannot start formatting. The path is the one thing wiki markup cannot hold exactly — the converter reports it:

1 backslash directly before a markup character was dropped: wiki markup cannot write that sequence.

The Cloud copy has no such limit: the same line pastes as text with C:\Temp\*.log intact in inline code.

Limits

What does not survive the conversion

How each Markdown feature is written for Confluence Cloud and for wiki markup
Markdown	Confluence Cloud copy	Wiki markup
Code block language	Carried in data-language; pick it in the snippet menu if Confluence shows another.	Written as {code:language=…} as typed; decorated info strings such as {r} are dropped.
Task lists	Action items when the whole list is tasks; a [ ] / [x] text prefix in mixed lists.	No task syntax: items keep a [ ] / [x] text prefix.
Footnotes	Superscript [label], with the note text where it was defined.	Superscript label, with the note text where it was defined.
Raw HTML	Shown as an HTML code block; comments removed, <br> kept as a line break.	Same, as {code:language=html}.
Table column alignment	Not carried over.	Not carried over.
Images	Linked by URL; relative paths such as images/a.png need uploading to the page.	Written as !url!; the same relative-path limit.
Quote inside a quote	Flattened into one quote.	Flattened into one {quote}.
Code or table inside a list item	Stays inside the item.	Written after the item; the list restarts below it.
Backslash before a markup character	Kept.	Dropped and reported: wiki markup cannot write it.

The converter lists the ones present in your text under the output, so you know what to check after pasting.

This converter is open source on GitHub and available on npm.

Questions

Markdown in Confluence: common questions

Can I switch a page back to the legacy editor to use Insert > Markup?

No. Atlassian retired the legacy editor in phases: new legacy pages stopped on January 21, 2026, and from April 1, 2026 every page opens in the cloud editor with no option to revert. The Insert > Markup dialog, which accepted wiki markup and Markdown, went with it. Pasting rich text is the built-in route that remains.

Can I keep editing the content as Markdown after pasting?

No. Confluence stores pasted content as its own rich text, so there is no Markdown source to reopen; Atlassian said the same of its old Markdown insert. Keep the .md file as your source and paste again after edits, or publish through the REST API from a script.

Is my Markdown uploaded anywhere?

No. Parsing and both conversions run in your browser with a JavaScript Markdown parser; nothing you type is sent to Kalcify. The only network requests the preview makes are for images your Markdown links to, fetched from their own hosts.

Sources

Atlassian Support — Confluence wiki markup syntax for the legacy editor
States that the Confluence Cloud editor does not support markup; wiki markup and the Markdown insert belonged to the legacy editor.
Atlassian Support — FAQ: Understanding legacy editor deprecation
Phase 3, April 1, 2026: the legacy editor is fully deprecated and pages can no longer be reverted to it.
Atlassian Support — Available markdown commands (Confluence Cloud)
The formatting the Cloud editor supports: headings 1–6, bold, italic, strikethrough, code, lists, quote, code snippet, divider, link, action item.
Atlassian Support — Add elements to a page or live doc
Code snippets auto-detect CSS, Go, HTML, Java, JavaScript, JSON, Python, Ruby, Rust, Shell, SQL and TypeScript.
Atlassian Documentation — Confluence Wiki Markup (Data Center 10.2)
Insert > Markup, the wiki syntax used here, where wiki markup is accepted, and the braced {_} form for formatting inside a word.
Atlassian Support — Unable to use some characters like backslash
A backslash escapes the next markup character, and a double backslash is a line break.
Atlassian Developer — Confluence Cloud REST API v2: Page
Page, blog post and comment bodies accept the representations storage, atlas_doc_format and wiki.
Atlassian — @atlaskit/adf-schema (editor document schema, Apache-2.0)
The paste parse rules this converter targets: pre[data-language] code blocks, th header cells, action-item markup, and monospaced pasted blocks becoming code snippets.

More free calculators and tools you might find useful

Compare Two ListsCompare two lists or two Excel/CSV columns and see items in both, only in A and only in B, plus a Yes/No match column. Two 100,000-line lists take ~0.15s.JSON FormatterBeautify or minify JSON with 2-space, 4-space or tab indentation, and get every syntax error reported with its exact line and column position.

Paste Markdown into Confluence Cloud

Why pasting raw Markdown turns into a code block

Markdown to Confluence wiki markup

How escaping works

Send the wiki markup through the Confluence REST API

Markdown in, Confluence out

A release checklist

Boundary case: characters that are markup in wiki syntax

What does not survive the conversion

Markdown in Confluence: common questions

Can I switch a page back to the legacy editor to use Insert > Markup?

Can I keep editing the content as Markdown after pasting?

Is my Markdown uploaded anywhere?

Sources

Related Tools