Privacy and data handling

The core value of Bookmark.new is to run AI summarization and embedding steps in your web browser, on your device, so that processing does not go to third-party cloud AI APIs.

We store your bookmark library on our servers (URL, title, scraped page text, summaries, tags, and embedding vectors) so you can sync across sessions and devices. AI inference stays on your device — we do not send page text to OpenAI, Anthropic, or similar services for summarization.

Your account, authentication session, and bookmark records live in our database. That includes the URL, title, optional summary text, tags, favicon and preview metadata, and embedding vectors we need for search—scoped to your account. We do not use your content to train public models; this is a private library product.

Paid plans are processed with Stripe. We receive billing and subscription state from Stripe (such as your subscription status and renewal dates) to unlock Pro features. We do not use Stripe to process the full text of the pages you bookmark.

Sign-in and related identity flows go through the auth stack configured for https://www.bookmark.new. Standard security practices apply (for example secure cookies in production, HTTPS, and the identity provider you choose where applicable).