Twitter Bookmark Limit: How Many Can You Save? (2026)

You tap the bookmark icon on a tweet without thinking. A thread on pricing strategy, a chart that made you stop scrolling, a tool you wanted to try later. The icon turns blue, you move on. A year passes. You have hundreds, maybe thousands of saves piled up. Then one day you go looking for the one that mattered, and the question lands.

Is there actually a limit on how many tweets you can bookmark on X? And if there is, what happens to the ones you saved first?

The answer is one of those rare cases where the official documentation, the API behavior, and the in-app experience all say slightly different things. X has never published a hard cap in its Help Center. The developer documentation pins the API at 800. The web app behaves like there is no ceiling on what is stored, then quietly stops showing you posts after roughly the same number. No wonder people online give you a different answer every time you ask.

This guide walks through what is actually true in 2026, why everyone has heard a different number, and what to do once your saved list gets large enough that the gap between “saved” and “findable” starts to bite.

Is there an official Twitter (X) bookmark limit?

The honest answer is that X does not publish one. The X Help Center lists what Premium gives you, and “more bookmarks” is not on the list. There is no setting that exposes a quota, no warning when you approach a number, and no upgrade prompt selling you out of a cap.

That leaves three places to look for a real answer: the X API documentation, the behavior of the bookmarks page in your browser, and the people quietly building third-party tools on top of both. They tell consistent stories, and none of them match the “no limit at all” framing on X’s marketing pages.

The picture that emerges, once you line up those three sources, is something like: you can keep tapping bookmark forever, X keeps recording the action, but the slice X gives you back is finite. The result feels like a soft cap that nobody at X will name out loud.

The 800-bookmark API limit, explained

The clearest cap is on the developer side. X’s own Bookmarks API documentation states plainly that the endpoint returns “800 of your most recent Bookmarked Posts.” Save number 801 does not get included. Save number 5,000 has not pushed any older save into an archive, it has just made the API stop a little further back from where you started.

This matters in real life because every serious third-party Twitter bookmark tool, the export utilities, the AI search tools, the read-later apps that sync with X, reads your saves through that same endpoint. They cannot return what X did not return to them. A paid bookmark manager that pulls “all your X bookmarks” is, in practice, pulling your last 800.

To make it worse, developers have reported intermittent pagination bugs on the X Developer Community throughout 2025 and 2026, where the API stops handing out the next_token cursor after only two or three pages of results. The realistic ceiling for some accounts has been more like 200 to 400 saves, not 800. If a tool ever told you it could not pull all your bookmarks, this is the underlying reason.

There is one exception. Bookmark folder contents, the new Premium API endpoint, have their own limits and have been even shakier. A separate thread on the developer forum reports the folder endpoint capping out at 20 items with no pagination. Folders, as we covered in X Bookmark Folders, look organized inside the app but are still difficult to read out programmatically.

The takeaway is simple. If your bookmark plan depends on a paid tool reading the X API, you are betting on the 800-most-recent slice. Anything older is gone, from the tool’s point of view, no matter how long you keep the subscription.

Why the web app feels like it has a 1,000-bookmark cap

The X website is a less precise picture. There is no documented cap on x.com itself, and accounts with very large libraries do appear to still hold them. A widely shared post from the team behind Twillot argues that X removed the old 800 limit years ago, and points at heavy users with more than 200,000 saves still intact. The underlying account, on that telling, will accept almost as many bookmarks as you want to give it.

What people report in practice, though, is a wall around 800 to 1,000. You scroll down your bookmarks page, you load more, you load more, and at some point the list stops giving you new posts. Older saves are technically still on your account, but they no longer surface in the timeline you see. This is closer to a display cap than a storage cap.

The split between “still saved” and “no longer shown” matters because users only experience the second one. If you cannot scroll to a save, the fact that it exists somewhere in X’s database does not help. From a user’s perspective, the bookmark might as well not exist. We dug into how this slow disappearance happens across X, Reddit and LinkedIn in Why Social Bookmarks Disappear.

The other thing the web app gives you is a single keyword search bar above your bookmark list. That search ignores anything past the visible window. So even if you remember the exact wording of an older save, typing it in returns nothing if the post lives below the cutoff. We covered that limitation, and the workarounds, in Search Twitter Bookmarks in 2026.

Does X Premium let you save more bookmarks?

This is the most common question once people understand the cap, and the answer is no.

X Premium adds two things to the bookmark experience: named folders to sort your saves into, and the keyword search bar over the main bookmarks list. Both are useful inside the visible window. Neither lifts the visible window itself. Premium reorganizes what you can already see. It does not bring back the older saves that have aged off.

X first introduced bookmark folders in 2023, when Social Media Today covered the rollout. Three years later the feature is still Premium-only, still useful for tidy savers, and still completely silent on the cap.

So if you are paying for Premium specifically to see more of your bookmark history, that is the wrong purchase. The right framing for Premium, if you save heavily, is that it lets you organize new saves more cleanly going forward. It does nothing for the year-old thread you remember but cannot find.

What happens to bookmarks past the visible cap

Once a bookmark drops out of the visible 800 to 1,000, here is what is true and what is not.

It is not deleted from your account. X is not removing the underlying save when a newer one comes in. The data appears to still live somewhere on your side of their database.

It is not retrievable through any official surface. The web app will not scroll to it, the keyword search will not find it, the X API will not return it past the 800-most-recent slice, and the official data export request, which we cover in detail in How to Export Twitter Bookmarks, pulls from the same recent window.

It is not rescued by Premium. The folders feature does not reach further back in history. The Premium API for folder contents is more limited than the public bookmarks endpoint, not less.

So practically, an aged-off save is a save that is no longer reachable by you, even if the row still sits in X’s database. The asymmetry is striking. The post still exists, the bookmark still exists, but the connection between you and the bookmark is broken on every interface you have access to.

This is the part nobody at X has ever had to defend in public. There is no clear policy, no announced behavior, no help center page that explains what happens to your oldest saves. Just a quiet ceiling that you only discover when you go looking for a specific post and it is gone.

Four ways to keep saves you cannot see anymore

If you already have a large bookmark library on X, you cannot recover the saves that have aged off. That door closed quietly some time ago, and X has given nobody a way to reopen it. Everything below is forward-looking: how to keep new saves in a place where the cap does not apply.

1. Request the official data export. From x.com/settings/your_data, you can ask X for a copy of your account history. Bookmarks are included. The export still pulls from the recent slice, so it does not recover older saves, but it gives you a one-time snapshot of what is currently visible. Run it today rather than next month, because the clock is already ticking on your oldest visible saves.

2. Use a tool that backs up bookmarks via the X API on a schedule. Several read-later apps, including Saverything, Dewey and TweetSmash, sync your bookmarks into their own database. New saves accumulate on their side, where the 800 ceiling does not apply once the post is captured. The catch is the same as for the export: anything older than that 800 window when you first connect is unreachable. Connect early.

3. Save to a separate tool at the moment you bookmark. A browser tool that watches your X activity and copies each save as you make it sidesteps the API entirely. The save lands somewhere outside X’s recent window the second it is made. Even if X aged it off the visible list a week later, your copy survives. This is the only method that genuinely beats the cap for everything you save after setup.

4. Be ruthless inside the visible window. If you do nothing else, work the recent slice. Tap unsave on posts you have already read, screenshot the chart or quote you actually needed, and let the rest go. The bookmark icon is a low-effort save signal, not a permanent shelf. Treating it as one keeps your visible 800 useful rather than crowded.

The first three are about scope. The fourth is about discipline. Most people who feel the bookmark cap acutely are mixing both problems, so it helps to separate them.

Twitter bookmark limit by method (and how to beat it)

A side-by-side view of where each method actually stands.

The pattern is the same one that runs through this whole topic. Every method that reads from X is limited by what X is willing to hand back, which is roughly the last 800 saves. The only method that escapes the ceiling is one that copies the save the moment you make it, before it has had any chance to age off.

The honest take on Twitter bookmark limits

Strip away the methods and one idea is left. X did not design bookmarks to be a library. Bookmarks behave like the back pocket of the timeline, a temporary holding area for posts you might want again soon. Anything that lives in the back pocket long enough quietly falls out, and the silence is on purpose. A library would imply an obligation to keep your data findable. A back pocket implies no such thing.

That framing also explains the half-answers from X itself. There is no published hard limit, because they have not committed to one. There is also no commitment to indefinite retrieval, because they have not committed to that either. The 800-bookmark API ceiling and the ~1,000-bookmark web scroll are the lived experience of an unstated policy. Until X changes that policy, every user with a large saved collection is going to keep running into it.

If you do not save heavily, none of this matters. The visible window covers everything you have ever bookmarked. The folders feature is a nice tidy-up, the search bar is a small extra. You will probably never hit the wall.

If you do save heavily, the conclusion is the same one we keep arriving at across the other guides in this series. The fix is not better organization inside the visible window. It is capturing each save the moment you make it, into a tool of your own, so the bookmark you reach for in two years is one X cannot quietly age off. The cheapest version of that is the official export, run today, then again every few months until you have moved your library somewhere durable.

Full disclosure: this is what we build, so weigh the next paragraph accordingly. ContextBolt is a Chrome extension that watches your saved posts on X, Reddit and LinkedIn and copies each save into a local knowledge base as you go. It stores the content, not just the link, so a save survives X aging it off the visible list. It tags every save by topic automatically, and searches by meaning, so you can look for “pricing strategy” and find a thread that talked about “what to charge” without ever using your exact words. We go deeper on that approach in Semantic Search for Bookmarks.

The honest scope: Basic is free and covers 150 saves with AI tagging and semantic search. Pro at $6/month lifts that to unlimited, adds encrypted cloud sync, and gives you an MCP endpoint so tools like Claude can read your saves directly. Because the extension captures going forward, it is the only method in this guide that is not capped by X’s 800-bookmark ceiling, for everything you save after you install it.

If your bookmark library is small, do nothing. If it is large enough that you have ever lost a saved post you wanted back, accept that the older saves are gone and start protecting the new ones today. The Twitter bookmark limit is not a number to memorize. It is a system X built around what they cared to keep visible, and the only response that actually works is to keep your own copy.