How to download your X archive (and what is inside)

#Quick answer

To download your X archive, open Settings and privacy, tap Account, then Your X data under Data and permissions, verify your identity with the code X sends to your email or phone, and tap Request data. X emails you when the export is ready and a Download archive button appears in the same settings screen. X's help page says preparing it "may take a few days", not the 24 hours most guides claim, and the file arrives as a zip containing an HTML viewer plus a data folder of your records.

Last updated: September 2026

#TL;DR

The request itself takes about ninety seconds. Everything interesting about this topic is on the other side of the wait: what X actually puts in the file, what the bundled viewer does badly, and what the export leaves out.

There is also a small myth to clear. Search results for this query are full of confident numbers, ready in 24 hours, link expires in 7 days, request limited to once a month. X publishes none of them. Its own page commits to "a few days" and says nothing about expiry, which is why the third-party numbers contradict each other. Plan around the vague promise rather than the specific one and you will not get caught out.

#How to request the export

In the mobile apps, per X's documented steps:

1. Tap your profile icon in the top menu, then Settings and privacy.
2. Tap Account.
3. Under Data and permissions, tap Your X data.
4. Verify your identity by tapping Send code to the email address or phone number on file. If you have neither on file, X redirects you to the Account information page first.
5. Enter the code.
6. Under Download your data, next to X, tap Request data.

On the web the shape is the same with one extra gate: click the more icon, choose Your account, click Download an archive of your data, enter your password and confirm, then verify with the code and hit Request data.

Two prerequisites from X's own note, and they are the cause of most failed downloads: confirm your email address before you request the archive, and open the download in the same browser where you are logged in to X.

When it is ready you get an email and, if you have the app installed, a push notification. You can download either from the email link or from the Download archive button that appears in the settings screen. The settings button is the more reliable of the two, since it does not depend on an emailed link still being valid.

#How long it takes, honestly

X's wording is: "It may take a few days for us to prepare the download of your X archive."

That is the whole commitment. In practice the wait scales with how much you have posted and how much media is attached, so a two-year-old account with a few thousand text posts is usually quick and a decade-old account with video is not. If nothing has arrived after several days, request again rather than waiting indefinitely, and check that the email on file is confirmed and current.

#What is actually in the zip

Unzip it and you get an HTML viewer, an assets folder, and a data folder. The data files use a .js extension but hold plain JSON with a one-line JavaScript assignment wrapped around the front, so stripping everything before the first [ turns any of them into a file your scripts can read directly.

X lists the contents of the export as: profile information, your posts, your Direct Messages, your Moments, your media (images, videos and GIFs attached to posts, DMs or Moments), a list of your followers, a list of accounts you follow, your address book, Lists you created or are a member of or follow, interest and demographic information X has inferred about you, and information about ads you have seen or engaged with.

Nobody opens the inferred-interest and ad-engagement files, and they are the two most worth ten minutes of your attention. They are the platform's own model of who it thinks you are, which is a more direct read on why your feed looks the way it does than any explanation of the X algorithm can give you.

#Five things the bundled viewer does badly

Every export ships with a Your archive.html viewer that renders your posts in a familiar interface. It is fine for browsing and poor for anything else. The maintainers of the widely used open-source archive parser catalogue the specific failures, and they match what you will hit within five minutes:

• Images open on the website, not locally. Click one to expand it and you are sent to x.com, which fails if you are offline or the account is gone, which is often the exact situation you built the archive for.
• Exported images are downscaled below what you originally uploaded.
• Posts sit in a complex JSON structure, so lifting them into a blog or a newsletter is a parsing job rather than a copy and paste.
• DMs are included without handles. Many of the user handles are missing from the export, so conversations arrive semi-anonymous.
• Every link is a t.co redirect, which hides the destination and stops working entirely if that shortener does.

None of this makes the archive useless. It makes the viewer a preview rather than a product, and it explains why a small ecosystem of parsers exists to convert the data folder into markdown, HTML or a spreadsheet.

#What the archive does not contain

• Anything you deleted. It is a snapshot of the account as it stands when the export runs.
• Other people's posts, including replies to you beyond what is needed for context.
• Analytics. Impression and engagement-rate history is not in here. That lives in your X analytics dashboard, and it is worth screenshotting before you need it.
• Live data. The moment the zip is built it starts going stale, which is the argument for exporting quarterly rather than once.

#Why this matters more in 2026

Until recently, pulling your own posting history programmatically was a free API call away. That ended when X moved the developer API to pay-per-use in February 2026 and closed the self-serve free tier, so every read of your own timeline now runs through a metered endpoint. Our breakdown of X API costs and access covers what that leaves for developers.

For an individual account the fallout is simple: the archive is now the free, complete, authenticated path to your own data. Not a scraper, not a third-party dashboard, and not the API. If you write on X and you would be upset to lose it, export quarterly and keep two copies. That takes ninety seconds a quarter and it is the cheapest insurance available on this platform. Anyone thinking about deleting their X account should treat the export as step zero, well before the deactivation flow.

If your interest is other people's public posts rather than your own history, the archive is the wrong instrument entirely and the constraints are completely different, which we covered in scraping X with Python.

#The best use of an old archive

Mine it. Load data/tweets.js into a spreadsheet, sort by likes, and read the top fifty. That is a portrait of what your audience actually responds to, built from evidence rather than from a template, and it usually contradicts what you think your best material is.

Then run the top posts again. Not word for word, but the same idea with a better hook, months later, to an audience that has mostly turned over. Republishing your own proven material is the highest-yield and least-used move in content on X.

The hard part is doing that consistently rather than in one nostalgic afternoon, which is the gap X-Autopilot fills: it drafts on-brand posts and replies through real Chrome on your own Mac, so a daily presence does not depend on you being free at the right hour. One honesty note, since your account is what is at stake: engaging through a browser tool is a gray area under X's automation rules rather than a sanctioned API route, and no tool removes account risk. Running locally on your own machine and IP is a smaller detection surface than cloud tools posting from shared pools, which is reduced risk, not zero.

Short version: request the export today, expect a few days rather than a day, keep the zip somewhere you will find it in five years, and open the data folder rather than the viewer.