Proxy detection
Recognising the kind of network a click arrived through — data centre, VPN, public proxy or TOR — as context about the request, not a verdict about the visitor.
Updated August 14, 2026
Proxy detection is working out what kind of network a click arrived through: a data centre or cloud host, a commercial VPN service, an open public proxy, a browser-based anonymiser, or a TOR exit node. It answers a question about the connection, not about the person using it, and those two questions are easy to confuse.
The work is a lookup rather than a judgement. Address ranges are classified in advance and distributed as a database, and a click's address either falls inside one of them or it does not. That makes the answer three-valued — a known network type, an ordinary address, or nothing at all because the range is not covered — and an unanswerable case should never be counted as a match, or an unavailable database turns into a wall.
Context is all the lookup gives you. Someone on a corporate VPN or a privacy-minded reader is an ordinary visitor; an address belonging to a hosting provider is rarely a phone in somebody's hand. Whether either fact matters is decided by the rule you write, not by the lookup itself.