About X Brain
X Brain is a web app that turns your X (formerly Twitter) liked posts into a private, searchable knowledge base. You upload the data archive that X gives you, and X Brain reads your likes, indexes them so you can search by meaning rather than exact words, sorts them into categories, pulls out the key takeaway from each one, and lets you export the whole thing as CSV, JSON, Markdown, or an Obsidian vault. It costs $19 once. There is no subscription.
Not to be confused with
Several unrelated organisations use a similar name, including xbrain.co.uk and xbrain.science, along with a cryptocurrency token that trades under the XBRAIN ticker. None of them are connected to this product in any way. X Brain is the software described on this page, and www.xbrain.live is its only website.
Why it exists
X has no search for your likes. Scroll back far enough and the timeline simply stops loading, which means years of things you deliberately saved become unreachable. Browser extensions try to solve this by scraping the page as you scroll, which breaks whenever the site changes and can never reach the posts that no longer load. X Brain takes the other route: it works from the official data archive, so it can see every like at once, including the oldest ones.
Who it is for
- People with hundreds or thousands of likes accumulated over years.
- Researchers who treat their likes as a reading list.
- Writers and builders who save references and cannot find them later.
- Obsidian and personal-knowledge-management users who want it all local.
What it cannot do
The honest limits, because they follow from the archive format rather than from the product:
- No author filtering.The archive’s
like.jsfile records only the post ID, its text, and a URL. It has no author field, so no tool can filter your likes by who wrote them without paid per-post X API lookups. - No “date I liked it”. The archive does not record it. Dates in X Brain are publication dates, recovered from the post ID, which is close enough to narrow things down by year but is not the same thing.
- Bookmarks are limited. Newer archives usually export bookmarks as bare IDs with no text, and text is what makes a post searchable. Likes are the reliable source.
- The archive takes time. X typically needs up to 24 hours to prepare your download. That wait is on their side, not ours.
Your data
Your archive is stored in Supabase with row-level security, so only your account can read your rows. If you supply your own AI provider key it is encrypted at rest. Nothing is sold or shared. You can export everything or delete all of it from Settings at any time. The full detail is in the privacy policy.
Contact
Questions, bugs, and refund requests all go to xbrainapp@outlook.com.
Ready to search your likes?
Upload your archive and browse it free. Pay only when you want the AI pipeline to run.
Get started