Privacy policy

Your email address

If you join the waitlist we store the email you enter, which page and form you used, and when you signed up. We use it for one purpose: sending you Relic emails, starting with a one-time welcome note that links you to the Relic beta, then occasional updates about Relic. We never sell it or share it for advertising. It is stored with our database provider (Supabase) and shared with our email delivery provider (Resend), which processes it on our behalf solely to send those emails. It is kept for as long as we send these updates, or until you ask us to remove it — reply to any email we send, or contact us below, and we'll delete it from both. If you create an account in the Relic app, its email address is added to this same list, so you may get these updates even if you never joined the waitlist; to stop them, reply "unsubscribe" to any email we send or contact us below, and we'll remove you from the list without affecting your account. Deleting your account does not remove your address from this list by itself, so ask us if you want both gone.

Cookies we set on this website

Two first-party cookies, both essential. "relic_consent" remembers the choice you make in the cookie panel and expires after six months. "relic_waitlist" is set only after you join the waitlist and holds a scrambled code derived from your email — not the address itself — so this browser can remind you that you're already on the list; it expires after one year. Neither cookie tracks you across sites, and neither is shared with anyone. If you choose Accept in the cookie panel, the Meta and Google pixels described below set cookies of their own.

Analytics and ad pixels on this website

Vercel Web Analytics loads only after you choose Accept in the cookie panel. It is cookie-free and aggregated — we see page views and country-level statistics, never a profile of you — and nothing follows you to other sites. If you reject, or simply never answer, it does not load at all.

Also only after you choose Accept, we load the Meta Pixel and the Google tag. They tell us which of our ads on Facebook, Instagram and Google bring people to this site and to the waitlist, and let us show our ads to people who have visited it. To do that they set cookies (such as "_fbp" and "_ga") and send Meta and Google the pages you view here, your IP address, and details of your browser and device, which those companies use under their own privacy policies — including to connect the visit to an account you hold with them. We never pass them your email address. If you reject, or never answer, neither loads; if you accept and later change your mind, reopen the cookie panel and choose Reject: the Meta Pixel stops at once, the Google tag stops using cookies, and neither loads on your next visit.

Our launch video

The launch video on the home page and on our video page plays from YouTube, which Google runs. We embed it in YouTube's privacy-enhanced mode (youtube-nocookie.com). The player loads with the page whether or not you accept cookies, so Google receives your IP address and details of your browser and device when you open those pages. Google says that in this mode YouTube does not store information about you unless the video plays. On the home page it starts playing, muted, when you scroll to it, so from then on YouTube may set cookies or store data on your device, as it may whenever you play the video yourself; Google uses what it collects under its own privacy policy. We get no data back about who watched.

Your account

You sign in with Google, or with an email address and password; sign-in is handled by Supabase Auth. We store your email address, your name if Google provides one, and the sign-in identity you used. If you sign up with email, we send you a confirmation email.

What we store when you use Relic

• Things you save: the link, the title, the readable text of the page (or the transcript, for a YouTube video), and page details such as the site name, author, favicon and publish date.
• Files you upload: PDFs and text or Markdown files, up to 50 MB each, kept in private storage that only your account can reach.
• What Relic generates from them: summaries, key takeaways, topics, categories and tags, and search indexes (text chunks and their numeric "embeddings").
• Chats: your questions and Relic's answers, including the passages it cited, until you delete the chat or your account.
• Digests and reading: your digest briefings, which items you marked, and how far you read (your progress, and the time spent reading an item).
• Settings: your display name, theme, layout, time zone and digest schedule.
• Activity history: a log of the changes you made, so you can undo them.
• Usage counts: how many saves, chat messages and import items you have used, for the free-beta limits.

Connected services: Notion and Google Drive

Only if you connect them. We store which pages or folders you chose and the name of the connected account (for Google Drive, that account's email address). Access tokens are stored encrypted in a secrets vault, not in our ordinary tables.

Google Drive is connected with read-only access, and we re-sync the items you chose about every 6 hours. Disconnecting a service revokes and deletes its tokens; items already imported stay in your library until you delete them.

Relic's use and transfer to any other app of information received from Google APIs will adhere to the Google API Services User Data Policy, including the Limited Use requirements.

Importing bookmarks

When you import a browser-bookmarks HTML file, we read it to show you its folder tree, and only the links you pick are saved. We do not keep the file itself. You can also import your bookmarks through the browser extension, described below: the same folder tree is shown, and again only the links you pick are saved.

The Relic browser extension

The extension asks your browser for four permissions, and one optional permission, each for one job:

• "activeTab" and "scripting" let it read the tab you are on, and only when you click Save, or when you open Relic on a YouTube playlist to import it.
• "tabs" lists your open tabs so you can pick which ones to save.
• "storage" keeps you signed in.
• "bookmarks" is optional. Chrome asks you for it only when you choose to import your bookmarks, and the extension works without it.

It requests no access to websites in general, and it does nothing in the background.

When you click Save, it reads the page as your browser already shows it and sends the page's link, title, readable text and details (favicon, author, site name, excerpt and publish date) to Relic. That can include pages you can see because you are logged in or subscribed — it saves what you were reading. Nothing is read from any page you did not click Save on, apart from a YouTube playlist you open Relic on to import, described below.

When the tab you save is a PDF, the extension reads the file in that tab, using your own browser session so that a PDF behind a login saves too, and uploads it to your private Relic storage, exactly as if you had uploaded it yourself.

When you save open tabs, the popup shows you your open tabs' titles and links. Only the tabs you tick are sent, and only their link and title. The list itself is not kept.

When you open Relic on a YouTube playlist to import it, the extension reads the playlist's video links and titles from that tab and shows them to you. Only the videos you tick are sent to Relic, which then fetches each one's transcript. The extension never changes your playlists.

When you import your bookmarks through the extension, it reads your bookmarks once, when you press Share on its own consent page, and passes them to the Relic app so you can pick folders and links. Only the title, link and folder name of the bookmarks you pick are saved. Nothing is read in the background or on a schedule, and the extension never creates, changes, moves or deletes a bookmark.

On your device it keeps your sign-in session (tokens), the progress of a save or import, and a one-time permit for a bookmarks import, in browser session storage that is cleared when the browser closes. To sign in, it opens app.relic-app.com, and that page passes your session to the extension.

The extension has no analytics, no tracking, no ads and no third-party scripts.

The use of information received from the Relic extension adheres to the Chrome Web Store User Data Policy, including the Limited Use requirements.

AI processing

Relic uses one AI provider, Amazon Bedrock, in AWS's EU regions. Anthropic's Claude models summarize and tag what you save, answer chat questions, and write your Digest. Amazon Titan creates the search embeddings that let Relic find things by meaning. Amazon Bedrock receives the content of your saved items, your chat messages together with the passages retrieved to answer them, your category names, and your search queries. Until October 22, 2026, search embeddings were created by Alibaba Cloud Model Studio (DashScope), international service, which received the text of saved items and search queries; it receives nothing from Relic since then.

Your content is never used to train AI models, by us or by our providers. It is sent strictly to produce the result you asked for.

Chat answers are based only on your own Library.

Fetching pages you save by link

When you save a link from the app (rather than with the extension), our servers fetch the page. If a site blocks us, we may ask a reading service (Jina Reader) or the Internet Archive's Wayback Machine for it, and for Reddit posts an archive service; YouTube transcripts may be fetched through a proxy network. Reddit posts are read through Reddit's own API or embed page, Instagram posts through Instagram's public embed page, and public YouTube playlists through YouTube's API. When you save an arXiv paper as a PDF, we also fetch arXiv's web version of that paper. These services receive only the public link or paper ID being saved — never your account or any other data.

Product analytics

We record first-party usage events to understand and improve Relic: things like signing in, saving, opening an item, searching, sending a chat message, and which app screen was viewed. Each event carries your account ID, the time, your device, browser and operating system type (not version strings or fingerprints), your country (from our network provider), and the app version.

We never record the titles, links or content of what you save, your chat text, your search words (only their length), or your IP address. The extension sends no analytics of its own.

Detailed events are deleted after 90 days, and per-day activity summaries after 13 months. Totals with no identifier attached, such as "accounts deleted today", are kept.

There is no third-party analytics in the app, and so no cookie banner: the app sets only the cookies needed to keep you signed in.

Cookies and storage in the app

• Sign-in cookies (your Supabase session), which are essential.
• On your device only: your theme, an unsent note draft, your reading position, and a per-visit ID used to group analytics events, which is cleared when you close the tab.

Who processes your data

These providers process data on our behalf to run Relic:

Where your data leaves the EEA or the UK, we rely on Standard Contractual Clauses or an equivalent safeguard. Until October 22, 2026, this covered our former embeddings provider, Alibaba Cloud Model Studio.

We do not sell personal data. The only data shared for advertising is what the Meta and Google pixels collect on this website, and only if you accept them in the cookie panel. A small number of Relic admins can see account email addresses and per-account activity details, to run the service: how many items you have saved and chat messages you have sent, your sessions and sign-ins, and your device type and country. They never see the titles, links or content of what you save, or your chats.

How long we keep your data, and deleting your account

• Your library, chats and settings are kept while your account exists.
• Deleting your account (Settings → Profile, where you type a confirmation phrase) erases your library, files, chats, search index, connected-service tokens, analytics and settings immediately.
• Encrypted backups are kept for up to 30 days, so deleted data fully disappears from them within about 30 days. Short-lived caches expire within 24 hours. Server logs, which contain an account ID but no content, are rotated by size.
• If you tell us why you are leaving, we keep the reasons you chose, any comment (500 characters at most), and the date. It is stored without your account ID or email address.
• Removing a connected service keeps the items you imported from it; delete them individually if you want them gone.

Exporting your library

Settings → Profile → Export creates a ZIP of your saved items (their text as Markdown, their details as JSON), your uploaded files, categories, digests and reading progress. The download link is valid for 1 hour. Your chats are included too, as chats.json: each chat session with its messages, citations and references.

Security

Your data is encrypted in transit (HTTPS everywhere) and at rest. Every request is checked against your account, and each account can only reach its own data. Backups are encrypted.

Your rights

Under the GDPR and similar laws you can ask what data we hold about you, have it corrected or deleted, get a copy of it to take elsewhere, object to how we use it, and withdraw any consent as easily as you gave it. Most of this you can do yourself: in the app, Settings lets you edit your profile, export your library and delete your account; on this website, reopen the cookie panel from the icon in the corner of the page at any time. For anything else, including your waitlist email, contact us below. We act on requests within 30 days. You also have the right to complain to a data protection authority, usually the one in the country where you live.

Children

Relic is not for anyone under 16, the same age limit set in our terms. We do not knowingly collect data from children; if you believe a child has signed up, write to [email protected] and we will delete their data.

Contact

Questions or requests about your data: email us at [email protected] — the same address every Relic email replies to, so "reply unsubscribe" and writing to us here reach exactly the same inbox. You can also find us on the Relic Discord; the link is in the navigation on the home page. Either way we'll take care of it.