How to export your X (Twitter) bookmarks — every method that works
No export button exists. These four routes do.
X has never shipped a bookmark export. Your saves sit in an endless scroll you can only read top-down, and the official data archive ("Download an archive of your data") does not include bookmarks. So getting them out means one of four routes. Here they all are, with the honest trade-offs — including the ones that cost money or cap how far back you can go.
01Manual copy
Open x.com/i/bookmarks, scroll, and copy each post's link into a doc or spreadsheet (share icon → "Copy link").
Free and safe, but it's minutes per bookmark. Fine for a couple dozen saves; hopeless for a pile.
02Capture from your own browser
Everything on your bookmarks page is already loaded in your browser — you're looking at your own data. A small userscript or extension can record each bookmark as you scroll the page yourself, then export the pile as JSON or CSV. No password sharing, no third-party server, works at whatever depth you're willing to scroll to.
- Install a userscript manager or extension you trust (read the code — it should write to local storage only, never phone home).
- Open your bookmarks page, logged in as yourself.
- Scroll at reading pace; the script records each post it sees.
- Export the captured list as a file on your machine.
This is the approach Rebrief's harvester takes: it records what your own browser shows you, on your own account, and stores it locally.
03The official X API
X's API moved to pay-per-use pricing in early 2026 (fractions of a cent per post read, no monthly minimum), which finally makes it affordable for personal use. But two hard limits matter for bookmarks:
- Depth cap: the bookmarks endpoint returns roughly your 800 most recent bookmarks — older saves are unreachable, at any price.
- Setup tax: you need a developer account, an app, and an OAuth flow before your first request.
Where it shines is ongoing sync: once you're set up, new bookmarks can be fetched hands-free on a schedule.
04Cloud bookmark tools
Services like Tweetsmash, Dewey, or Circleboom connect to your account via OAuth and pull bookmarks into their dashboard, usually for a monthly fee.
Convenient, and fine if you're comfortable with the trade: your saves get copied to their servers, live behind their account system, and remain yours only as long as the subscription and the service survive. They inherit the same ~800-bookmark API depth limit for backfill.
| Method | Cost | How far back | Where data ends up | Effort |
|---|---|---|---|---|
| Manual copy | Free | Unlimited (you scroll) | Your doc | Very high |
| Your own browser | Free | Unlimited (you scroll) | Your machine | Low, one-time setup |
| Official API | Pay-per-use + setup | ~800 most recent | Wherever you point it | Medium (dev account) |
| Cloud tools | Subscription | ~800 most recent | Their servers | Lowest |
Depth caps and pricing are as of July 2026 — X changes these without much notice. This guide describes reading your own bookmarks from your own logged-in account; automating other people's content or bulk collection is a different thing entirely, and against X's terms.
Exported. Now what?
A JSON file of 600 bookmarks is still a pile — just a different pile. Rebrief turns it into a weekly briefing: one takeaway per save, sorted by why you saved it, computed on your machine. No account, no cloud. The backstory: why bookmarks are a graveyard, and what we did about it.
Get early access →FAQ
Does X's official data archive include bookmarks?
No. The "Download an archive of your data" export covers posts, likes, DMs, and profile data — bookmarks are not in it. That's why the methods above exist.
Can the X API export all my bookmarks?
No — the bookmarks endpoint only reaches back through roughly your 800 most recent saves, regardless of what you pay. Anything older is only reachable by scrolling your own bookmarks page in a browser.
Is capturing from my own browser allowed?
You're reading your own bookmarks, on your own account, at your own pace — the same data X is already sending to your screen. Keep it personal-scale and local: don't share credentials with third parties, don't collect other people's content in bulk, and read the current X terms yourself, since they change.
What format should I export to?
JSON if a tool will process it (it keeps author, date, and text intact), CSV if you want a spreadsheet. Whatever you pick, keep the post URL — everything else can be re-derived from it.