Rummage — semantic search for your bookmarks

Rummage makes the bookmarks you already have searchable by what they are about, not by whether you remember the exact title. For example, search "python visualization" and find the plotting library you bookmarked two years ago under a name you have long forgotten. WHAT MAKES IT DIFFERENT: There is nothing to re-save and nothing to migrate. Rummage is not another read-it-later app asking you to start a new habit; it works on the bookmark folders already sitting in your browser. HOW IT WORKS: 1. You pick which bookmark folders Rummage may read. Anything you leave unchecked is never opened. 2. It reads the pages in those folders in the background, newest first, and builds a search index. 3. You search. Results are ranked by how closely they match your meaning, with a score you can hover to understand. WHAT STAYS PRIVATE: The search index lives in your browser. There is no account, no sign-up, and no sync server. Your bookmarks are not stored on our side. To make a page searchable by meaning, its text has to be processed by a language model, so Rummage sends the title, folder name and roughly the first 1,500 characters of each page through our own server to an embedding model. The result comes back and is stored locally; nothing is retained on our side. Your bookmark URLs are never sent at all. Rummage never uses your logged-in sessions or cookies. That is a deliberate trade-off: pages behind a login or paywall cannot be read, so they are indexed by title and folder only and clearly labelled "title only" in results. NO API KEY NEEDED: Rummage covers the processing cost, on the free version as well as the paid one. There is nothing to sign up for and no key to paste. Full privacy policy: https://get-rummage.com/privacy Terms and refunds: https://get-rummage.com/terms Support: support@get-rummage.com