More unique IPs than any network on the planet, in the world's largest residential pool. That's Shifter.More unique IPs than any network, in the world's largest pool. That's Shifter.

See the numbers

Pool overlap

The same addresses, sold to you twice.

Every network is tested in the same window, so we can intersect the exact exit IPs each one handed us. Where two pools share addresses, one provider is selling inventory you can already reach through the other.

18 networks, each matched against every other
Two pools, one shared block
Live
Pool A
Pool B
In both
  • Same window
  • Exact exit addresses
  • Matched pair by pair
  • Resale flagged
01 · The market

Every pool, counted once

Add every network's pool together, then count each shared address once. The gap is inventory sold more than once.

Sum of pools

71.8M

Every network's unique IPs, added up

Counted twice

43.6M

Addresses sold by more than one network

Distinct IPs

28.2M

What is left once each address counts once

Added up, these networks sell 71.8M IP addresses. 43.6M of them (60.7%) are copies of an address another provider also sells. Counted once each, that leaves 28.2M different addresses.

Resale signatures

12 of 18 networks have most of their pool inside one larger network: what buying wholesale looks like from the outside.

  • Decodo logo Decodo Resale signature

    Most of Decodo's observed exits also appear in Oxylabs, and most of Oxylabs's appear here too. The two pools are largely the same addresses.

    Decodo inside Oxylabs 81.7%
    Oxylabs inside Decodo 81.9%

    A shared block of 5.74M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • Proxy-Seller logo Proxy-Seller Resale signature

    Most of Proxy-Seller's observed exits also appear in Infatica, and most of Infatica's appear here too. The two pools are largely the same addresses.

    Proxy-Seller inside Infatica 78.4%
    Infatica inside Proxy-Seller 70.0%

    A shared block of 2.96M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • Infatica logo Infatica Resale signature

    Most of Infatica's observed exits also appear in Shifter, but little of Shifter's appear here. The relationship runs one way.

    Infatica inside Shifter 74.7%
    Shifter inside Infatica 38.9%

    A shared block of 3.60M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • FleetProxy logo FleetProxy Resale signature

    Most of FleetProxy's observed exits also appear in Shifter, but little of Shifter's appear here. The relationship runs one way.

    FleetProxy inside Shifter 73.5%
    Shifter inside FleetProxy 45.8%

    A shared block of 5.26M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • Massive logo Massive Resale signature

    Most of Massive's observed exits also appear in NodeMaven, but little of NodeMaven's appear here. The relationship runs one way.

    Massive inside NodeMaven 71.4%
    NodeMaven inside Massive 26.7%

    A shared block of 1.42M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • Byteful logo Byteful Resale signature

    Most of Byteful's observed exits also appear in Shifter, but little of Shifter's appear here. The relationship runs one way.

    Byteful inside Shifter 65.4%
    Shifter inside Byteful 27.3%

    A shared block of 2.54M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • Webshare logo Webshare Resale signature

    Most of Webshare's observed exits also appear in Oxylabs, but little of Oxylabs's appear here. The relationship runs one way.

    Webshare inside Oxylabs 63.8%
    Oxylabs inside Webshare 38.3%

    A shared block of 2.40M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • ProxyEmpire logo ProxyEmpire Resale signature

    Most of ProxyEmpire's observed exits also appear in Shifter, but little of Shifter's appear here. The relationship runs one way.

    ProxyEmpire inside Shifter 63.4%
    Shifter inside ProxyEmpire 38.5%

    A shared block of 3.83M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • SOAX logo SOAX Resale signature

    Most of SOAX's observed exits also appear in Shifter, but little of Shifter's appear here. The relationship runs one way.

    SOAX inside Shifter 58.5%
    Shifter inside SOAX 8.1%

    A shared block of 727K addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • Evomi logo Evomi Resale signature

    Most of Evomi's observed exits also appear in Shifter, but little of Shifter's appear here. The relationship runs one way.

    Evomi inside Shifter 57.3%
    Shifter inside Evomi 6.1%

    A shared block of 497K addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • Proxy-Cheap logo Proxy-Cheap Resale signature

    Most of Proxy-Cheap's observed exits also appear in Shifter, but little of Shifter's appear here. The relationship runs one way.

    Proxy-Cheap inside Shifter 54.8%
    Shifter inside Proxy-Cheap 4.1%

    A shared block of 308K addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

  • NodeMaven logo NodeMaven Resale signature

    Most of NodeMaven's observed exits also appear in ProxyEmpire, but little of ProxyEmpire's appear here. The relationship runs one way.

    NodeMaven inside ProxyEmpire 51.3%
    ProxyEmpire inside NodeMaven 49.9%

    A shared block of 2.75M addresses. Resale is not misconduct, but paying a markup for it, or running both expecting independence, is a costly mistake.

Next · 02 Exclusivity How much of each pool is its own?
02 · Exclusivity

What you can only get here

The share of each pool that appears in no other network we measure. Higher means you are buying something you cannot get elsewhere.

Pool exclusivity per network
Network Pool size Exclusive IPs Shared Largest overlaps
Shifter Resale signal 8.95M 2.61M
70.8% FleetProxy logo 45.8% Infatica logo 38.9% ProxyEmpire logo 38.5%
Geonode Mixed sourcing 3.29M 1.42M
56.9% Rayobyte logo 25.2% Shifter logo 23.6% NodeMaven logo 19.5%
IPRoyal Mixed sourcing 2.89M 1.23M
57.3% Webshare logo 25.7% Decodo logo 24.5% Oxylabs logo 24.4%
Decodo Resale signal 7.03M 776K
89.0% Oxylabs logo 81.7% Webshare logo 38.0% Shifter logo 16.0%
Oxylabs Resale signal 7.17M 765K
89.3% Decodo logo 81.9% Webshare logo 38.3% Shifter logo 16.1%
Webshare Resale signal 3.76M 752K
80.0% Oxylabs logo 63.8% Decodo logo 63.5% IPRoyal logo 17.1%
Rayobyte Resale signal 2.10M 653K
68.9% Geonode logo 30.7% NodeMaven logo 27.4% Byteful logo 25.3%
FleetProxy Resale signal 7.15M 393K
94.5% Shifter logo 73.5% ProxyEmpire logo 60.9% Infatica logo 55.9%
ProxyEmpire Resale signal 6.04M 384K
93.6% Shifter logo 63.4% FleetProxy logo 62.4% NodeMaven logo 49.9%
NodeMaven Resale signal 5.36M 339K
93.7% ProxyEmpire logo 51.3% Shifter logo 50.7% Infatica logo 48.6%
Massive Resale signal 1.98M 331K
83.3% NodeMaven logo 71.4% ProxyEmpire logo 45.3% Infatica logo 14.0%
DataImpulse Resale signal 881K 298K
66.2% NodeMaven logo 27.6% IPRoyal logo 27.3% Shifter logo 25.0%
Byteful Resale signal 3.88M 281K
92.8% Shifter logo 65.4% Infatica logo 64.7% Proxy-Seller logo 60.6%
SOAX Resale signal 1.24M 189K
84.8% Shifter logo 58.5% FleetProxy logo 53.6% NodeMaven logo 52.8%
Infatica Resale signal 4.82M 142K
97.1% Shifter logo 74.7% Proxy-Seller logo 70.0% FleetProxy logo 67.0%
Proxy-Seller Resale signal 3.78M 134K
96.4% Infatica logo 78.4% Shifter logo 76.5% FleetProxy logo 65.0%
Evomi Resale signal 867K 114K
86.8% Shifter logo 57.3% FleetProxy logo 53.1% ProxyEmpire logo 47.6%
Proxy-Cheap Resale signal 562K 34K
93.9% Shifter logo 54.8% Infatica logo 52.1% Proxy-Seller logo 51.1%
Read the exact rules

Exclusive is the share of a network's observed exit addresses that appeared in no other network in this set. Shared is the rest. Both come from one matched run, so they read the same in every window.

Exclusive IPs

Each network's unique IPs, estimated the same way as everywhere in Atlas, multiplied by the share of its observed exits found on no other network: the addresses you can't reach through any other network in this set. The table is sorted by them.

Largest overlaps

Each logo is one of the three networks this pool overlaps with most, beside the share of this pool also seen there. The same shared block is a different share of each pool, because pool sizes differ: a small pool can sit almost entirely inside a large one while being a sliver of it.

Next · 03 By network Which networks share addresses?
03 · By network

Where each pool also shows up

One bar per network. Green is the part of its pool found nowhere else, and every other colour is a share also seen in another network. Hover a segment for the exact figure.

18 of 18 networks share part of their pool

Next · 04 How it works How is overlap measured, and what does it not prove?
04 · How it works

How overlap is matched

Exact exit addresses, from the same requests in the same window, intersected pair by pair.

  1. 01

    Same test schedule

    Every residential proxy network is tested on the same schedule against the same targets, so results line up.

  2. 02

    Real exit IPs

    Each request logs the exit IP the target actually saw, not the number a proxy provider's dashboard reports.

  3. 03

    Shared IP pools

    Exit IPs are compared for every pair of networks. Shared counts match; percentages differ with pool size.

  4. 04

    Resold proxy pools

    A residential pool sitting mostly inside one larger network is flagged as likely resold proxy inventory.

What overlap does not prove

Four limits to keep in mind when reading the figures

  • Resale is not fraud

    Many resellers add real value in routing or support. The risk is a hidden premium.

  • Overlap is a floor

    We only intersect what we observed, so pools may share more than we saw, never less.

  • ASNs are not IPs

    Two networks on one ISP look alike by ASN yet share no addresses. Only exact IPs count.

  • The trend matters most

    A network growing its own pool shows falling overlap, so the trend says more than one run.

FAQ

Frequently asked questions

Common questions about shared residential proxy pools and how Atlas measures them.

How much of one network's pool also shows up in another network's pool. Atlas measures it on exact exit IP addresses, so a shared address means both networks handed out the same IP to our tests.

Every network is tested on the same schedule, and each request records the exit address the target saw. Those address sets are intersected for every pair. The shared count is the same both ways, but the percentages differ, because pool sizes differ.

Much of that pool is also sold through another network. A pool that sits mostly inside one larger network is flagged as likely resold inventory. That does not make the two services equal: a reseller can still differ on routing, support and price.

Not necessarily. Overlap shows shared exit addresses, not who supplies whom. Since Atlas can only intersect the addresses it observed, two pools may share more than the figures show, never less.

Get started

Check the overlap before you pay twice

Shifter's pool is matched against every network in Atlas by the same rules. Order with a 3-day money-back guarantee and check its exits against the provider you use today.

3-day money-back guarantee. Cancel anytime.