Last updated 17 November 2026. Questions about this policy go to admin@instavision.co.

1. Information we collect

When you create an account we collect the details you provide (your email address, and your name and profile picture if you sign in with Google), the searches you run — the descriptions you type, including your conversations with the search assistant — and their results, the feedback you give us (messages, your ratings of the accounts a search found, and your answer when we ask whether a search found what you needed), your credit balance and payments (card details stay with our payment provider; we never see full card numbers), and basic usage data needed to run and improve the service — such as the days you use it and when you open a result, export it or save accounts to a list.

2. Public profile data in search results

Search results contain information that account holders have made publicly available on their Instagram profiles: the username, name, bio, follower and post counts, public posts and captions, and contact details they chose to list publicly. We obtain it from third-party providers of publicly available web data and check every account against the search ourselves. We do not collect anything from private accounts or from behind a login.

We process this data only to show the user who ran a search which accounts match it and why, on the basis of our legitimate interest in helping our users find businesses and creators who have chosen to be publicly discoverable (Art. 6(1)(f) GDPR). Some of it may count as personal data. If your account appears in our results and you would rather it did not, see “Your choices” below — you can object at any time.

InstaVision is not affiliated with Instagram or Meta.

3. How we use data

We use your information to provide and improve the service, to keep it secure, and to communicate about your account. We do not sell personal information, and we do not use search results for anything but showing them to the user who ran the search.

4. AI assistants

When you connect an AI assistant (for example Claude or ChatGPT) to InstaVision, it signs in with your permission and sends us the requests you make through it: search descriptions, search ids, filters and the handles you add to your blocklist. We use them only to run those requests on your account. Your requests and their results are also processed by that assistant's provider (for example Anthropic or OpenAI) under its own terms. We count, per day, which assistant software connects and which tools are called — the software's name, never your account or the content of a request. You can disconnect an assistant at any time in Settings → API keys.

5. Who receives data

We use these providers to run the service: Supabase (database, sign-in and file storage), Vercel (hosting), Stripe (payments), Resend (email), Cloudflare (the human check on the sign-in page), Google (if you sign in with Google), Anthropic (AI models that read your search description and the public profiles we check), providers of publicly available web data (to collect public Instagram profile data for your searches), and the analytics and error-reporting tools listed in section 8. Each receives only what it needs for its part of the service.

6. How long we keep it

Your account details, searches, results and the emails we sent you are kept while your account exists. PDF export links expire after one hour. Sign-up checks are kept as described in section 8 (keyed hashes of the IP address and browser fingerprint, for 30 days). Public profile data in results is kept while the account of the user who ran the search exists, or until the profile's owner asks us to remove it.

7. Your choices

You can ask for a copy of your data, or for your account and its searches to be deleted, by writing to admin@instavision.co from the address you sign in with.

If your Instagram account appears in our results and you would rather it did not, write to admin@instavision.co with your username. We remove it from every user's results and saved lists and stop showing it in future searches.

8. Analytics and cookies

We use the tools below to see how the site is used and to fix what breaks. Some of them set identifiers in your browser, so we don't call them anonymous.

Yandex Metrica (Yandex) counts page views, clicks and form submissions (never what you type). It does not record sessions. It does not run in our internal admin area, or on a page whose address carries an invitation link, or an email address in its query. The page addresses we report as page views keep only campaign tags (Yandex's own ad-click id included) and a few page settings, with record ids replaced by placeholders; click reports carry the page's address as it is. Where a link to another website leads is not reported. Yandex Metrica sets its own cookies and browser-storage entries, among them the cookies _ym_uid (identifies a browser, 1 year), _ym_d (the date of your first visit, 1 year) and _ym_isad (an ad-blocker check, 20 hours). Yandex lists its cookies, with their lifetimes, at yandex.com/support/metrica/general/cookie-usage.html.

Vercel Web Analytics counts page views without cookies: Vercel recognises a visit by a hash of the request and discards it after 24 hours. The addresses of our pages it receives are cleaned the same way. Vercel Speed Insights measures how fast our pages load, without cookies, and receives page addresses cleaned the same way.

Google Analytics (Google) counts page views, scrolling and the start of form filling (never what you type). The page addresses it receives are cleaned the same way as for Yandex Metrica, except that Google's own ad-click ids are kept; where a link to another website leads is not reported. It does not run where Yandex Metrica does not. It sets the cookies _ga (identifies a browser, 2 years) and _ga_<id> (keeps the visit, 2 years).

Microsoft Clarity (Microsoft) records how pages are used — clicks, scrolling and mouse movement — to build heatmaps and session recordings. Everything you type into a field is masked, and inside your account every piece of text on the page is masked too; only the public pages (home, pricing, use cases, legal) are recorded with their text. It receives each page's address as it is, so it does not run where Yandex Metrica does not. It sets the cookies _clck (identifies a browser, 1 year) and _clsk (joins the pages of one visit, 1 day); visitors from the EEA, the UK and Switzerland get no Clarity cookies.

Sentry receives a report when something breaks. It also records a sample of sessions, and the sessions that hit an error, with all text masked and all images and media blocked; link and page addresses are recorded as they are.

Our own cookies. These four can be read only by our server:

  • iv_did: a random identifier for this browser, used to spot repeat sign-ups for the free credits. 1 year.
  • iv_ft: how you first arrived (see below). 1 year.
  • iv_login_next: the page to return to after you sign in. 24 hours.
  • iv_login_email: the email address you entered on the sign-in page, so the next page can show it. 24 hours.

These can also be read by scripts on the page:

  • iv_goal: which way you signed up (Google, an emailed link, a code or an invitation), for our analytics. 10 minutes.
  • iv_fp: a fingerprint of this browser — a one-way hash of settings such as its screen size, languages, time zone and how it draws graphics, not the settings themselves — used to spot repeat sign-ups for the free credits. Set on the sign-in page, and only on desktop browsers. 30 days.
  • iv_search_draft: the search you typed on the home page, carried into your first search. Deleted once it is shown, or after 24 hours.
  • Supabase session cookies (sb-…): keep you signed in. Up to 400 days.

The site also keeps a few entries in your browser's storage, for example so that a starter prompt is sent only once; Sentry keeps its own while it records a session.

The first time you arrive from a campaign link, an ad or another website, a first-party cookie (iv_ft) remembers for up to a year the campaign tags, the domain of the referring website (not the page), the page you landed on (with any ids or tokens removed) and the time you arrived. For an ad click it keeps only which kind of click identifier the link carried (for example gclid), never the identifier itself, and it drops anything that looks like an email address. If you sign up with an emailed link, these details may be copied into your new account's sign-in record when the link is sent, so they are not lost if you open the link on another device, and we remove them from that record once your sign-up has been recorded.

When you sign upwe keep the first-visit details above, a hash of the browser identifier, a normalized copy of your email address and its domain (and whether that domain is a throwaway-mail service, which we check by looking up the domain's public DNS records), and whether the sign-up looks like a repeat of an earlier account. For 30 days we also keep a keyed hash of the IP address your sign-in link was opened from and of the browser fingerprint above — never the address or the settings themselves — and we note whether another account signed in or out on our site in the few minutes before yours was created. We use this to protect the free credits from abuse: a sign-up that looks like a repeat does not get them and may be closed. If that happens to you by mistake, email us and we will restore it.

9. Contact

Questions or data requests? Email admin@instavision.co.

Questions or data requests?Email us