Guide · tested with Docmost 0.96.0 on 3 October 2026
How to move from Confluence to Docmost
Docmost is an open source wiki that feels a lot like Confluence. Its one-click Confluence importer needs a paid license. The free Community edition can still take a Confluence export through its zip import, but you'll be rebuilding the page tree by hand. Here's exactly what happens, so you can decide which route fits.
The short answer
- Small space, a few dozen pages: the free zip import plus an hour of tidying works. Follow the steps below.
- Hundreds of pages, or a deep page tree: the free import puts every page at the top level with file names as titles. Use Docmost's paid Confluence importer, or have someone convert the export first.
- Page-level permissions matter to you: Docmost only has those in its paid edition, whichever way you import. Access is otherwise set per space.
How we tested
We ran the official docmost/docmost Docker image (version 0.96.0, the latest on 3 October 2026) with Postgres and Redis, as in Docmost's install guide. We imported two Confluence HTML space exports, one shaped like Data Center and one like Cloud, with nested pages, panels, code, expand blocks, status labels, task lists, images, a PDF and links between pages. The exports were made up for testing, so try your own space before you rely on any of this.
Step 1: Export your Confluence space as HTML
In Confluence, export the space as HTML. You'll get a zip with one HTML file per page, an index.html, and folders for attachments and images. Our step-by-step export guide covers Cloud and Data Center. Leave the zip as it is.
Step 2: Check Docmost's size limit
Docmost accepts import zips up to 200 MB by default. Raise FILE_IMPORT_SIZE_LIMIT in its environment if your export is bigger, and FILE_UPLOAD_SIZE_LIMIT (50 MB by default) if single attachments are. Both are in Docmost's environment variable docs.
Check this before you import. When we sent a zip over the limit through the API, the upload was accepted but cut off at exactly 200 MB, and the import then failed with "file is truncated". Splitting a large space into several exports is another option we haven't tested, and it leaves you more tree to rebuild.
Step 3: Import the zip
- Open the space you want the pages in.
- In the sidebar, next to Pages, choose … then Import pages.
- Under Import zip file, choose Upload file and pick the Confluence zip.
The Confluence button in the same dialog is greyed out on the free edition, with the note "Available with a paid license". The zip import is the free route. Our small test spaces took about a second each.
What you get
| In Confluence | After the free zip import |
|---|---|
| Page tree | Flattened. Every page lands at the top level of the space, sorted A to Z by file name. |
| Page titles | File names, such as Local-Setup_65541, or just the page ID (65542) when the file name is a number. |
| Space home page | An extra page made from index.html, with the space details and a list of pages. |
| Page text, headings and lists | Kept. Each page also starts with Confluence's breadcrumbs, title and "Created by" line, and ends with an "Attachments" list and a "Document generated by Confluence" footer. |
| Code blocks | Kept, but the language is lost and a code title becomes a bold line above the block. |
| Info, note, tip and warning panels | Plain paragraphs. The panel title becomes an ordinary line of text. |
| Expand blocks | Always open, the title as an ordinary line. |
| Status labels | Plain text. |
| Task lists | Plain bullet lists. Tick boxes and done states are lost. |
| Images | Uploaded and shown, though images that sat inside a sentence move to their own line. |
| Attachments | Uploaded when a page links to them (the export's attachments list does), named by Confluence ID (65546.pdf rather than setup-guide.pdf). PDFs show in a viewer. |
| Confluence emoticons | Broken images. |
| Links between pages | Data Center style exports: plain links point to the new pages, but links to a heading (…html#Section) break, and full Confluence addresses stay as they were. Cloud style exports: most links still point to your old Confluence site. |
| Comments, page history, authors and dates | Page history isn't in an HTML export. Authors and dates survive only as the "Created by" line of text on each page. |
| Page restrictions | Not carried. Page-level permissions are a paid Docmost feature. |
Step 4: Tidy up
- Rebuild the tree. Open
index.html's page from the import to see the original list, then drag pages under their parents in the sidebar. - Rename pages. Each page's real title is in its first heading, after the space name.
- Delete the Confluence leftovers at the top and bottom of each page: breadcrumbs, the "Created by" line, the attachments list and the footer.
- Turn notes back into callouts and task lists back into to-dos with Docmost's
/menu. - Fix links. Search for your old Confluence address and for
.html#to find links that still point to the old site or to a heading. - Keep the export zip. It's your snapshot of the old space.
Budget roughly a minute or two per page. That's fine for 30 pages and painful for 500.
Other routes
- Docmost's paid edition includes a Confluence importer that Docmost says keeps the page tree, callouts, toggles, internal links, attachments and draw.io diagrams. We couldn't test it without a license. See Docmost's import docs.
- Convert the export first. Docmost's zip import keeps nesting and page order when the zip follows the layout of Docmost's own exports: folders that match the page tree, plus a
docmost-metadata.jsonfile. A script that rewrites the Confluence export into that layout fixes the tree and titles. Macros still need converting separately. - Have someone do it. That's what we do, below.
Moving to Docmost with Tablewide
New Tablewide workspaces run on Docmost, hosted for you at one flat monthly price with unlimited users, and founding teams get their Confluence space moved for free. Our own importer rebuilds the page tree in Confluence's order with the real page titles, and turns panels into callouts, status labels into status pills, task lists into to-dos and expand blocks into toggles. Code keeps its language, and attachments, images and links between pages carry over. In our tests on 13 sample exports (2,120 pages), every page landed in the right place.
Some things don't carry over even with us. Links to a heading open the page at the top. Labels don't carry over yet. Pages show the importer as author. Page restrictions aren't carried; access is set per space. You check the result before anyone switches.
Docmost is a trademark of its owners and Confluence is a trademark of Atlassian. We're not affiliated with either.