Proxy chaining
Routing traffic through more than one proxy in sequence, so each hop only knows its neighbours, not the whole path.
Proxy chaining sends your traffic through two or more proxies one after another. Each link in the chain sees only the hop before and the hop after it, so no single proxy knows both who you are and where you are ultimately going. It is the layering idea behind onion routing, applied deliberately.
The appeal is defence in depth: compromising one proxy does not unmask the whole path. The cost is latency, because every extra hop adds a round trip, and complexity, because a chain has more places to break or leak than a single proxy.
For most commercial proxy work a single well-chosen residential exit is enough, and chaining is overkill that mostly buys you slowness. It matters when the threat model genuinely requires that no one link can see the full route.
The latency cost is not a rounding error, and it is the reason chaining rarely survives contact with a real workload. Every hop adds a full round trip, and hops are usually in different places, so a chain routing through two continents can add several hundred milliseconds to every single request. On a job of any size that is the difference between an afternoon and a week.
It also multiplies the failure surface rather than adding to it. Two hops means two things that can be down, two sets of credentials, two places a DNS leak or a misconfiguration can occur, and a debugging problem where a failure gives you no indication of which link caused it. Reliability is the product of the links, not the average.
The honest assessment is that chaining answers a threat model most commercial users do not have. If your concern is that a site should not see your address, one good exit achieves that completely, and a second hop adds nothing the first did not already provide. If your concern is that no single operator anywhere in the path should be able to correlate you with your destination, that is a genuinely different requirement, and it is the one chaining exists for.
Combining a VPN with a proxy is the common practical version, and it does have a specific use: it hides proxy use from your own network while presenting the proxy's address to the destination. That is a real reason. Stacking hops for a vague sense of extra safety is not, and it costs you speed for nothing.
Frequently asked questions
Does chaining proxies make me more anonymous?
Against a specific threat, yes: no single operator in the chain sees both who you are and where you are going. Against the ordinary concern, that a website should not learn your address, it adds nothing, because one good exit already achieves that entirely. Most people asking have the second concern and pay the cost of solving the first.
How much slower is a proxy chain?
Roughly the sum of every hop, and hops are usually far apart. Each one adds a full round trip, so a chain crossing two continents can add several hundred milliseconds to every request. On a large job that compounds into a completely different timescale, which is why chaining is rare in production scraping.
Can I chain a VPN and a proxy?
Yes, and it is the most common practical combination. Traffic goes through the VPN first and then the proxy, so your local network sees only the VPN and the destination sees the proxy. The genuine use is concealing proxy use from the network you are on. It does not make you meaningfully harder to identify at the destination.
Is proxy chaining the same as Tor?
The layering idea is the same and the engineering is not. Tor routes through multiple relays with encryption layered so each relay knows only its neighbours, run by volunteers on a public network, and it is slow by design. A manual proxy chain is a handful of hops you arranged, with none of Tor's cryptographic guarantees about what each hop can learn.
Back to the full glossary.