A face search that tells you
what it can't do.

What actually happens

1. 1

   You upload a photo

   The file is checked by its actual header rather than its extension, and passed to the index provider's detector to find a face. This step costs nothing and happens before you are asked for money — a photo with no usable face is rejected here, because selling someone a search that cannot succeed is indefensible.

2. 2

   The face becomes a signature

   The detected face is cropped, squared and converted into a numerical signature describing the geometry between its features. Your original photo is not stored: it is the query, not the haystack.

3. 3

   You choose a plan

   Nothing has been searched at this point. Each search costs us real money at the index whether or not you buy anything, which is why the paid step is the search itself rather than the results.

4. 4

   The search runs

   One search from your allowance is deducted, then the signature is compared against the full index. It takes roughly twenty-five seconds and the progress you see on screen is the real thing, not an animation on a timer.

5. 5

   You get the results

   Every match, ranked, with a 0–100 confidence score, the platform it came from and a direct link to the page. Results are kept for thirty days so you can return to a report you paid for, then deleted.

What we refuse to do

This is the useful part of an about page. Every item below is checkable against the product rather than a promise about our character.

• We don't show you fabricated matches

  It is common in this category to show a wall of blurred results with invented handles and scores before payment. Those searches have not run. We show you what a report looks like, clearly labelled as a sample, and nothing else — because charging for evidence that does not exist is fraud, not marketing.

• We don't publish an accuracy percentage

  There is no honest way to produce one. Move the threshold and you can manufacture any figure you like.

  The long version →

• We don't invent statistics or reviews

  The figures on this site are read live from the engine or aggregated from searches actually run. There are no testimonials, because we would have to write them.

• We don't build pages about named individuals

  Mass-produced pages targeting people's names are the standard growth tactic here. They rank for a while, they are miserable for the people named in them, and they are what gets sites in this category deindexed. We don't make them.

• We don't help with deanonymisation or outing

  Checking whether an account contacting you is real is legitimate. Attaching a real name to a pseudonymous person who is not doing anything to you is harassment. So is searching to find out whether someone sells adult content. Both are grounds for termination.

• We don't hold your photo

  The upload is deleted once the search completes. Locked previews are genuinely destroyed server-side rather than CSS-blurred, so the browser never receives an image the paywall claims to be hiding.

Where the index comes from

We do not crawl the web ourselves. ARGUS is built on a third-party face index of publicly accessible images, and we pay for every search that runs against it. That relationship is why the search is the paid step, and it is also the limit on what we can see: exactly what a signed-out visitor could reach, and nothing behind a login.

Who runs this

Operator details are not configured yet. Set NEXT_PUBLIC_COMPANY_NAME, NEXT_PUBLIC_COMPANY_LOCATION and NEXT_PUBLIC_CONTACT_EMAIL to show them here. This block is deliberately empty rather than filled with a plausible-looking company — search engines and customers both check, and an invented one is worse than none.

If you want out

If your face appears in results and you want it gone, we will de-index it. It is free, it does not require you to have been a customer, and there is no upsell attached to it.