It estimates a region and shows it with a search radius and a confidence score - an educated guess from visual clues like signage, vegetation, and architecture, not a pinpoint. Treat it as a lead to verify, not a certainty. It works best on outdoor scenes with recognizable landmarks or signage.

No. DeepSearch analyzes the SCENE - location, street clues, vehicles, authenticity - and never runs facial recognition or identifies the person. That line does not move.

It varies with the photo. Distinctive outdoor scenes with clear signage or landmarks estimate well; plain interiors, close-ups, and nature shots have far less to go on. Every estimate ships with a confidence and a radius so you can see how much to trust it.

Public guides and illustrative examples are free to read. Running a search requires paid access. Plans include a usage allowance, and coverage depends on the available sources.

Yes. DeepSearch only analyzes the photo you provide and public, openly available information. It is not a background-check or FCRA consumer report, so results can't be used for employment, credit, housing, insurance, or tenant decisions.

If your photo carries GPS metadata, it's read on our servers to check authenticity and then stripped before the image is stored or analyzed - it's never shared, and it's never shown back to you as a 'result'.

It's a Beta signal. When a photo carries provenance metadata (like a camera or an AI generator tag) the check is strong; otherwise it's a weak heuristic, and we label it as such. It does not detect face-swaps - that needs forensic analysis a vision model can't do.