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

Atlas methodology

Cannot reproduce it? Then it is marketing.

Every figure in Atlas comes from one collector, one written target suite and one published scoring rule. Nothing is hand-tuned, and Shifter is measured the same way as everyone else.

46,293,653 requests across 18 networks and 8 targets in the last 24 hours146,064,199 requests across 18 networks and 8 targets in the last 4 days146,064,199 requests across 18 networks and 8 targets in the last 4 days
One batch, through the pipeline
Live
Sent
Passed
Failed
Published
  • Same tests for all
  • Published formula
  • No paid ranks
  • Open data
The method

One request, start to score

Every number in Atlas comes from the same path. These are its six stops.

  1. 01 What we test products and targets
  2. 02 Send and judge one request, one verdict
  3. 03 What it tells us Every figure, 3 windows
  4. 04 One score 8 parts, 1 grade
  5. 05 The limits what it cannot say
  6. 06 Check it do not trust us
How every score is calculated

Every request counts for every metric, and every time slot weighs the same. Every sub-score is a ratio to the best network in Atlas (100 = the leader). Target scoring: 80% success + 20% speed under load, each against the leader. A network that is down is data, not a gap. Overall scores require complete, fresh inputs; missing values remain unavailable.

The rules behind this score, in one file. Its SHA-256 is the version shown above. View the rules (JSON)

Verify it yourself: the command prints the version.

curl -s https://atlas.shifter.io/v1/methodology/88131d128c2bd29076cfbe3a83921cf29d5a95a03b34c3b226716e698eecccb1.json | shasum -a 256

88131d128c2bd29076cfbe3a83921cf29d5a95a03b34c3b226716e698eecccb1

Methodology version: 88131d128c2b

01 · What we test

Every residential product, tested as sold

Atlas measures residential proxies: rotating and sticky, each tested as customers buy it.

Rotating residential

A new exit IP per request or per short session, drawn from a shared pool.

Live IPs
Pool tracking
Overlap
Sticky TTL

Sticky residential

A residential exit held for an advertised TTL, for multi-step sessions.

Live IPs
Pool tracking
Overlap
Sticky TTL
Read the exact rules

Rotating residential

109 tries a day per target

All countries with usable coverage · 237 countries tested

This is the product most buyers are actually choosing between, and the one where claims are least checkable. It gets the full suite: every target, continuous exit-IP sampling, churn tracking and the cross-network overlap audit.

Sticky residential

109 tries a day per target

All countries with usable coverage

Same suite as rotating, plus TTL verification: we hold a session for its full advertised duration and record whether the exit IP survived. Advertised TTLs are ceilings, and the gap is never disclosed.

Next · 02 Send and judge How is one request sent and judged?
02 · Send and judge

How one request gets a verdict

A 200 is not a success. Each response is checked against a written rule for its target.

  1. [01]

    Acquire

    One live request per target through the proxy gateway.

  2. [02]

    Classify

    Exit IP mapped to country, city, ASN and IP type.

  3. [03]

    Assert

    Must pass a written check. HTTP 200 is not success.

    • → Pass

      The response met the written condition for that target. · success

    • → Challenge

      A CAPTCHA came back as HTTP 200 instead of the page. · failure

    • → Block

      The target refused: 403, 429 or a reset at the edge. · failure

    • → Timeout

      No usable response within the target's time budget. · failure

  4. [04]

    Aggregate

    Success, latency and IP pool, scored against the leader.

  5. [05]

    Diff and publish

    Changes logged, full dataset published as open JSON.

Read the exact rules

The five steps

01 Acquire

A collector node opens a session against the provider gateway with our own credentials, requests one target, and records the full response: status, headers, timing, body assertions and the exit IP the target saw.

02 Classify

The exit IP is looked up against public registry data for country, state, city and ASN, then classified as eyeball, hosting, mobile or business. Nothing is taken from the provider dashboard.

03 Assert

A 200 response is not a success. Each target has a written pass condition, such as a parsed element, a JSON shape or a token, and a response that fails it is recorded as a challenge, a hard block or a timeout.

04 Aggregate

  • Every request counts, whether it came from a census, a pulse or a load wave.
  • Results are grouped into time slots as long as the pulse interval, and each slot weighs the same.
  • Latency histograms are merged rather than averaged.
  • Each sub-score is the network's value as a share of the best network in Atlas.
  • No manual adjustment happens at any point.

05 Diff and publish

The run is diffed against the previous one, changes above the reporting threshold are recorded, and the whole dataset is published as a single JSON file alongside Atlas.

The four outcomes

Only a pass is a success; the other three are the failures that vendor dashboards report as traffic served.

Pass

success

The response satisfied the written condition for that target: the parsed element, the JSON shape or the token is present in the body we got back.

Challenge

failure

A CAPTCHA or interstitial was served instead of the page. It arrives as HTTP 200, which is how success figures based on status codes end up looking almost perfect.

Block

failure

The target refused outright: 403, 429, a block page, or a connection reset at the edge before any content was served. Still counted as traffic.

Timeout

failure

No usable response inside the budget for that target, including errors raised by the proxy gateway itself rather than by the site.

Recorded for every request

Address
exitIp
the address the target saw, not the gateway
country
from registry data at collection time
state
region or equivalent, same lookup
city
the city that registry assigned the address
asn
looked up from registry data at collection time
asnName
the operator name attached to that ASN
kind
eyeball, hosting, mobile or business
Response
status
and the assertion verdict beside it
ttfb / total
connect, first byte, last byte
bytes
payload size, which sets the real cost
Session
sessionId
to measure how long an exit was held
target
which site this request was aimed at
vantage
which of the eight collector regions issued it
Next · 03 What it tells us What does each request tell us about a network?
03 · What it tells us

Every figure, four questions

Each request adds to every figure below, grouped by the question a buyer asks.

{
// IP pool · What a network really hands out.
"ipPool": {
  • "uniqueIps": int, // Unique IPs
  • "churnRate": pct, // Pool churn
  • "recurringShare": pct, // Repeat IPs
  • "requestsPerIp": float, // Requests per exit IP
  • "stickinessHoldRate": pct, // Sticky session hold rate
  • "sharedShare": pct, // Pool overlap
  • "claimShare": pct, // Pool vs claim
},
// Performance · How it behaves on real targets.
"performance": {
  • "successScore": score, // Success score
  • "latencyP50": ms, // Latency
  • "challengeScore": score, // Challenge score
  • "uptimePct": pct, // Availability
  • "downtimeMinutes": minutes, // Downtime and incidents
},
// Authenticity · Whether the pool is genuine.
"authenticity": {
  • "datacenterAsnShare": pct, // DC ASNs
  • "datacenterIps": int, // Datacenter IPs
  • "residentialAuthenticity": pct, // Authenticity
  • "blacklistHitRate": pct, // Blocklist presence
  • "leakChecks": { dns, webrtc, … }, // Leak checks
},
// Coverage and cost · Where it works and what it costs.
"coverageCost": {
  • "countries": int, // Country / state / city / ASN totals
  • "listRate": usd, // Value and $ / GB
  • "costPer1kAllTargets": usd, // Cost per 1,000 successes
  • "attemptsPerSuccess": float, // Attempts per success
}
}

Every figure can be read over three time ranges

  • 24h Last 24 hours

    The freshest view: today's defences on every target.

  • 7d Last 7 days

    Adds repeat IPs, which take more than one day to see.

  • 30d Last 30 days

    The range the Atlas Score and ranking use.

New data every hour · 46,293,653 requests in the last 24 hours · 146,064,199 requests in the last 4 days · 146,064,199 requests in the last 4 days

Read the exact rules

All 21 definitions, and why nobody publishes them

IP pool

Unique IPs

Distinct exit addresses actually observed answering our requests.

Advertised pool size describes inventory. It is unverifiable by construction.

Pool churn

Share of exits in the window never seen in any previous window.

Churn determines whether your session survives, and nobody publishes it.

Repeat IPs

Share of distinct exits in the 7-day or 30-day window that appeared on two or more days.

A rotating residential pool should keep this low. A high figure is inventory that behaves like static IPs.

Requests per exit IP

Mean consecutive requests one address served before the gateway rotated us away.

It is the difference between a sticky product and a rotating one, regardless of marketing.

Sticky session hold rate

Share of sticky sessions that survived their advertised TTL.

Advertised TTLs are ceilings, not guarantees, and the gap is never disclosed.

Pool overlap

Share of the network's unique IPs also seen in another network's pool; the rest is its own (exclusiveShare). The network holding most of it is named as the likely upstream (likelyUpstream).

Reselling is normal. Paying a markup for the same addresses without knowing it is not.

Pool vs claim

Unique IPs seen in the window as a share of the pool size the network advertises.

An advertised pool is a claim. This is how much of it actually shows up.

Performance

Success score

A 1 to 10 score per target, from the share of requests that met the written pass condition. The best network on each target reads 10; the others keep their measured distance.

Every request counts (census, pulse and load); each time slot weighs the same, so a busy census hour cannot drown out a quiet night. A network's overall score is the mean of its target scores, each weighted by the target's difficulty, 1 to 5.

A single success figure is usually quoted against an undefended endpoint, where nearly every request succeeds on every network.

Latency

Time from request start to last byte on normal requests (the hourly pulse), through the proxy, as p50, p90 and p99, on the plain-HTTP target only. The latency under heavy load is shown beside it.

  • HTTPS adds CONNECT and TLS round trips (about twice as slow), so other targets show success and blocks but no latency.
  • Each runner region picks its fastest country as its reference, once a day from the previous day's light-load p50.
  • In every pulse slot that country gets the 100-request pulse and, 60 seconds later, a wave at census concurrency.
  • Load latency ÷ pulse latency is the region's slow-down factor, applied to the light-load latency of every country tested from that region.
  • Light-load latency (pulses only) is shown next to it. Speed is scored on the under-load p50.

Vendors quote connect time on an idle connection, which excludes both the part you wait for and the slow-down you hit at volume.

Challenge score

A 1 to 10 score per target, ranked on the share of requests served a CAPTCHA or interstitial: fewer challenges score higher. A challenge counts as a failure.

A challenge returns HTTP 200, so any success figure based on status codes silently counts it as a win.

Availability

Outage time per country tested, so an outage in one of two countries costs half.

1 − downtime ÷ (window × countries). An outage opens when a pulse finds the network down and closes with the next healthy pulse; a census the network spoiled is rerun once it is back.

Proxy outages are almost always partial, so binary uptime always reads 100%.

Downtime and incidents

Minutes of measured outage in the window, and the outages opened in it (incidentCount), each with its severity, reach and length.

Status pages report what the provider admits. These are the outages the collector hit.

Authenticity

DC ASNs

Share of unique exit IPs whose IP database usage type is hosting or datacenter, among IPs with a usage type, on networks sold as residential.

This is the single most misrepresented property in the market.

Datacenter IPs

Unique exit IPs on hosting or datacenter networks: the count behind the DC ASN share.

A share can hide a large number on a large pool.

Authenticity

Share of unique exit IPs that the IP database places in the country the requests asked for, among IPs with a known country.

A pool sold by country is only worth what it delivers in that country.

Blocklist presence

Share of sampled exits appearing on public reputation lists. Shown where measured; not part of the score.

It predicts failures you will blame on your own code.

Leak checks

DNS, WebRTC, proxy headers, TLS fingerprint and IPv6 fallback, asserted from the target side.

These are sold as features and almost never verified.

Coverage and cost

Country / state / city / ASN totals

Distinct countries, first-level subdivisions, cities and autonomous systems among the observed exits. Higher is better on every count.

Advertised "195 countries" is a catalogue claim. These totals are what the collector actually saw.

Value and $ / GB

$ / GB is the price of the first plan of 200 GB or more (the largest plan when none is that big), with the range up to the entry plan.

Value is the network's place on cost per 1,000 successes among the priced networks: 100 for the cheapest, falling evenly to the last.

The entry rate is what a trial costs, not what a buyer at volume pays.

Cost per 1,000 successes

Price normalised by the share of requests that passed and by payload, across all eight targets.

On each target: (1,000 × measured data per request) × lowest $ / GB ÷ the share of requests that passed, so failed requests are paid for. The eight targets are averaged with equal weight.

Rate cards price bandwidth. Nobody buys bandwidth.

Attempts per success

Requests one success takes over the window: requests ÷ successes.

Every failed attempt is billed bandwidth. This is the retry tax a rate card leaves out.

How each figure is read in a window

Every figure can be read over the last 24 hours, the last 7 days, or the last 30 days. The collector writes a bucket every 5 minutes, close to real time on the pools. The longer windows are that same stream looking further back.

Requests are spread evenly across the window and across targets, so a network cannot look better by being fast at 04:00. Session and rotation settings are left at each provider's defaults; we measure the product as sold, not a tuned configuration.

Figure by figure

Success and latency

All windows
  • Every request counts.
  • The window is cut into time slots (the least common multiple of the pulse interval and the five-minute bucket: 30 minutes, 10 in staging) and every slot weighs the same.
  • Each target's pass share is the mean of its slot shares before it is scored 1 to 10, and latency merges each slot's light-load histogram at equal weight.
  • Latency is the HTTP target's only; latency under load is the mean of its slot values.
  • Pulses and load waves that ran while a census of the same network was running are left out of light-load latency, slot success and the slow-down factor, but still count as requests, for uptime and as slots.

Requests and bandwidth

All windows

Counted across the selected window.

Pool size

All windows

Distinct exits seen in that window: the last 24 hours, the last 7 days, or the last 30 days.

Repeat IPs

7d · 30d

Share of unique exits seen on two or more days, over the last 7 or 30 days. Not shown on 24 hours, which cannot show a multi-day repeat.

Cost per delivered result

All windows

List price stays the same; success changes with the window, so cost per delivered result is restated for each one.

Availability and incidents

All windows

Outage time over the selected window: 1 − downtime ÷ (window × countries tested). A network that is down is data, not a gap: its slot counts as filled and down, lowers uptime and opens an incident. Last 24 hours, last 7 days, or last 30 days.

Per-target results

All windows

Each target's scores follow the selected window, as every other figure does.

The window control on the Overview switches every figure at once. Where a figure is only measured over 24 hours, it stays on 24h rather than moving.

Cross-network pool overlap: how resold pools are found

The measurement we have not seen published anywhere else, and the one most likely to change a purchasing decision.

Because every network is tested in the same window with the same requests, we hold the actual exit addresses each one handed us. Intersecting those sets answers a question the market never addresses: how much of a pool is genuinely that provider's inventory, and how much is the same addresses you can buy elsewhere.

Shared counts are symmetric between a pair. The two percentages are not, because pool sizes differ, and the asymmetry is the finding. A small pool mostly contained inside a much larger one, where the reverse is barely true, is resold inventory.

shared(A, B) = |exits(A) ∩ exits(B)|
shareOfOurs = shared(A, B) ÷ |exits(A)|
shareOfTheirs = shared(A, B) ÷ |exits(B)|
flagged as resale when shareOfOurs > 45% and shareOfOurs > 3 × shareOfTheirs

Conditions that make it valid

01

Both networks tested concurrently. Pools rotate constantly, so sampling two providers on different days produces a low overlap that means nothing.

02

The exit address is taken from what the target observed, never from a provider dashboard.

03

Only address-level matches count. Two networks sourcing from the same ISP will share an ASN while sharing no addresses, and conflating the two would be a serious error.

04

Overlap is reported as a floor. We intersect what we saw; two pools may share more than that, never less.

Next · 04 One score How do the figures become one score?
04 · One score

From measurements to one grade

Each part is compared with the best network in Atlas, weighted, and added up.

Atlas Score = Σ(sub-score × weight) ÷ Σ(weight)

Each part is yours divided by the best network of your type. The leader scores 100. 100 × yours ÷ best

  • Reliability 22%
  • Speed 14%
  • Value 14%
  • Coverage 12%
  • Availability 10%
  • DC ASNs 10%
  • Authenticity 10%
  • Diversity 8%

0 to 100

The weighted mean of the eight parts, shown with a grade.

These are the default weights. The leaderboard lets you re-weight the mix for your own workload.

Read the exact rules

How the score is built: the four rules and the eight sub-scores

Every part is measured against the best network in Atlas. There are no fixed targets to hit.

more is better: success, availability, unique IPs, locations, clean share, authenticity
sub-score = 100 × yours ÷ leader
less is better: latency p50, cost per 1,000 successes
sub-score = 100 × leader ÷ yours
both clamped to 0–100, then
Atlas Score = Σ(sub-score × weight) ÷ Σ(weight)

The four rules

01 Like against like

Every network in Atlas is a residential proxy network, tested on the same targets with the same requests, so every score compares like with like.

02 A ratio to the leader

On every part, the best network in Atlas scores 100 and everyone else scores their share of it. Half the leader's pool is 50; twice the leader's latency is 50.

03 Fair to new networks

Counts (IPs, countries, states, cities, ASNs) are daily counts averaged over the days a network has data, so one day of testing compares fairly with thirty. Rates such as success, latency, availability and cost use the whole window.

04 Always the last 30 days

The headline score covers the last 30 days, or fewer for a network that is newer. Leaders are found again every time scores are calculated, a few minutes after each test batch, so the 100 moves when a network gets better.

Notes on the parts

  • Diversity mixes four ratios: 30% countries, 20% states, 25% cities and 25% ASNs, each against the leader on that count.
  • DC ASNs uses the clean share, 1 − the share of unique exit IPs whose IP database usage type is hosting or datacenter.
  • Authenticity is 100 × the share of unique exit IPs that the IP database places in the country the requests asked for, set against the leader.
  • Value is 100 × the cheapest network's lowest $ / GB ÷ this network's lowest $ / GB.
  • A network that is the only one with a value for a part is its own leader and scores 100. If the leader's value is zero, nobody can be a share of it and the part scores 0; a latency or cost of zero scores 100.
  • The weights are the defaults above and do not change with the leaders.

The eight sub-scores

Each one reads a single measurement and sets it against the best network in Atlas, so 100 always means the leader. The tag says which direction is better.

How it performs
01 22%

Reliability

Share of requests that passed, each time slot weighted equally, per target

Higher is better 100 × yours ÷ leader
02 14%

Speed

Median latency of normal requests (the hourly pulse), p50, on the HTTP target

Lower is better 100 × leader ÷ yours

Each light-load (100-request) latency × the slow-down its runner region measured at census concurrency in its fastest country. HTTPS and other targets carry no latency.

03 10%

Availability

Capacity-weighted availability over the window

Higher is better 100 × yours ÷ leader
What the pool is
04 12%

Coverage

Unique exit IPs answering requests, averaged per day

Higher is better 100 × yours ÷ leader

Live unique exits, not the advertised pool size.

05 8%

Diversity

Countries, states, cities and ASNs seen, averaged per day

Higher is better 0.30·r(countries) + 0.20·r(states) + 0.25·r(cities) + 0.25·r(ASNs)

r(·) = 100 × yours ÷ leader on that count.

06 10%

DC ASNs

Clean share: 1 − share of exit IPs whose IP database usage type is hosting or datacenter

Higher is better 100 × yours ÷ leader

Hosting IPs in a residential pool lower this part.

Label and cost
07 10%

Authenticity

100 × share of exit IPs located (IP database) in the country the requests asked for

Higher is better 100 × yours ÷ leader

Per unique exit IP; IPs without a known country are left out.

08 14%

Value

Lowest list price per GB of the pool tested

Lower is better 100 × leader ÷ yours

The worked example: today's ranking

Each figure over the last 30 days, with the sub-score it earns below, coloured by its place

Network Success weight 22 Latency p50 weight 14 Availability weight 10 Unique IPs weight 12 Cities · ASNs weight 8 DC ASNs weight 10 Authenticity weight 10 Cost / 1k weight 14 Atlas Score
1 Shifter 9.4/10 99.3 486 ms 71.2 100.00% 100.0 8,948,662 100.0 103,632 · 24,797 98.7 0.6% 99.5 100.0% 100.0 $4.43 88.9 94.1 A+
2 Oxylabs 9.9/10 99.9 480 ms 72.1 99.95% 100.0 7,167,058 62.6 101,067 · 22,478 96.4 1.3% 98.7 100.0% 100.0 $10.03 44.4 83.4 A
3 Decodo 10.0/10 100.0 504 ms 68.7 99.97% 100.0 7,032,104 63.6 100,995 · 22,414 96.7 1.3% 98.7 100.0% 100.0 $11.38 33.3 81.5 A
4 FleetProxy 6.8/10 97.2 919 ms 37.6 99.82% 99.8 7,150,983 59.2 81,755 · 18,319 84.8 0.7% 99.3 100.0% 100.0 $9.07 72.2 80.6 A
5 Geonode 9.5/10 99.1 810 ms 42.7 99.95% 100.0 3,286,685 25.3 55,566 · 17,508 78.3 0.0% 100.0 100.0% 100.0 $3.30 94.4 80.3 A
6 DataImpulse 6.3/10 98.9 471 ms 73.5 99.73% 99.7 880,984 10.7 43,269 · 9,932 63.9 0.0% 100.0 100.0% 100.0 $4.45 83.3 80.1 A
7 ProxyEmpire 7.6/10 97.5 799 ms 43.3 99.78% 99.8 6,036,035 54.9 92,380 · 20,256 91.2 0.5% 99.5 100.0% 100.0 $9.54 61.1 79.9 B+
8 Webshare 8.8/10 99.6 770 ms 44.9 100.00% 100.0 3,759,734 40.1 84,722 · 18,356 87.6 1.0% 99.0 100.0% 100.0 $9.44 66.7 79.3 B+
9 NodeMaven 7.3/10 97.9 512 ms 67.6 99.32% 99.3 5,356,729 51.9 97,252 · 21,120 94.2 0.6% 99.4 100.0% 100.0 $16.01 22.2 77.7 B+
10 SOAX 3.1/10 99.3 418 ms 82.8 100.00% 100.0 1,244,359 12.4 47,634 · 13,493 71.0 1.3% 98.7 100.0% 100.0 $9.75 50.0 77.5 B+
11 Proxy-Seller 6.4/10 97.1 753 ms 46.0 100.00% 100.0 3,778,210 41.1 79,584 · 17,490 84.9 0.7% 99.3 100.0% 100.0 $9.69 55.6 77.2 B+
12 Byteful 9.7/10 99.8 517 ms 66.9 100.00% 100.0 3,884,045 36.3 65,731 · 12,042 57.9 1.3% 98.7 100.0% 100.0 $10.84 38.9 75.6 B+
13 Infatica 6.5/10 97.6 709 ms 48.8 99.91% 99.9 4,818,619 47.1 80,869 · 18,151 86.0 0.8% 99.3 100.0% 100.0 $13.91 27.8 74.6 B+
14 Massive 7.0/10 99.1 346 ms 100.0 99.95% 100.0 1,984,735 17.9 67,363 · 12,363 59.2 0.1% 100.0 100.0% 100.0 $23.52 11.1 74.2 B+
15 Evomi 8.6/10 99.0 384 ms 90.1 99.80% 99.8 867,174 10.7 40,903 · 10,855 63.4 0.0% 100.0 100.0% 100.0 $20.07 16.7 73.1 B+
16 Rayobyte 1.0/10 77.5 888 ms 39.0 99.94% 99.9 2,101,773 16.5 60,854 · 17,869 79.5 2.4% 97.6 100.0% 100.0 $7.49 77.8 71.5 B
17 Proxy-Cheap 1.0/10 75.6 1.02 s 33.9 92.34% 92.3 562,185 9.2 35,760 · 7,408 60.6 0.6% 99.4 100.0% 100.0 $2.73 100.0 70.5 B
18 IPRoyal 7.2/10 98.5 1.08 s 32.1 99.92% 99.9 2,892,443 26.0 66,362 · 17,149 81.7 1.6% 98.4 100.0% 100.0 $25.40 5.6 66.4 B
  • Read one cell: Shifter's latency is 486 ms against Massive's 346 ms, so its speed is 100 × 346 ÷ 486 = 71.2.
  • Every sub-score and every Atlas Score here is the one Atlas publishes, read from the same data as every other page.
  • Value is by place on cost per 1,000 successes: 100 for the cheapest, falling evenly to the last.
Next · 05 The limits What can this score not tell you?
05 · The limits

What the score cannot tell you

Six honest gaps, so you know when to test for yourself.

  • A collector node is one vantage point

    Latency is measured from our nodes, so yours will differ.

  • Pool size is a floor, not a total

    We count the IPs that answered, not the inventory a provider claims.

  • Plan tier changes results

    We buy a mid-tier plan. A different contract can perform differently.

  • Targets defend themselves differently over time

    A success score describes one week, not the network forever.

  • We do not test at abusive volume

    Low, rate-limited traffic says little about very large scraping loads.

  • Overlap does not reach the Atlas Score

    Pool overlap is published in full but never folded into the score.

Read the exact rules

Stating limitations is not a disclaimer, it is part of the method. Every one of these is a reason your numbers may legitimately differ from ours.

A collector node is one vantage point

Latency is measured from our nodes, not yours. A network that looks slow from Frankfurt may be fast from São Paulo. Region is recorded per request and the collector runs from eight regions, but your own numbers will differ and that is expected.

Pool size is a floor, not a total

We report distinct exit IPs we actually saw in a window. A provider claiming 90 million addresses is not necessarily lying; they are describing inventory, while we describe what answered. Both numbers are useful and only one can be checked.

Plan tier changes results

Some networks route paying tiers through better inventory. We buy a mid-tier plan and say so. If your contract differs, test it yourself and compare.

Targets defend themselves differently over time

Anti-bot vendors ship changes weekly. A success score is a measurement of one week, not a property of a network. Trends matter more than single numbers.

We do not test at abusive volume

Request rates are deliberately low and rate-limited per target. This keeps the test ethical, and it also means we do not measure how a network behaves under the load of a large scraping operation.

Overlap does not reach the Atlas Score

Pool overlap changes decisions about redundancy and effective pool size rather than describing quality, so it is published in full and folded into nothing. An Atlas Score that absorbed it would obscure both.

Left out on purpose

What is deliberately left out

Customer reviews are published beside the data and reach the Atlas Score nowhere.

Why reviews stay out of the Atlas Score

A written review captures things no measurement can reach: whether support answers, whether billing is honest, whether the documentation is usable. That is genuinely valuable, and it is why reviews are published next to the measurements.

It is also unreproducible, which is the one property the Atlas Score has to have. Averaging an opinion into a measurement produces a number that is neither, so the two sit side by side and are never mixed. A five-star network with a bad score is telling you something specific, and blending them would erase it.

How a review is verified

A verified review requires proof of purchase: a billing document showing the provider name and a month. We check those two facts, publish the month only, and delete the file.

Unverified reviews are published and labelled, and excluded from the verified average.

Next · 06 Check it How do you check any of this?
06 · Check it

Do not take our word for it

The data and the formula are public. Three ways to check.

  1. Check our arithmetic

    01

    Download the exact dataset Atlas was built from and recompute any number on it.

  2. Recompute the score

    02

    Apply the published formula to the dataset and you get the same ranking.

  3. Run your own test

    03

    Run our open-source tests on your own account with the AI coding agent you already use, then compare.

The strongest is the third: test a network on your own account and compare your figures with ours. If the shapes match and only the absolute numbers differ, the method is sound and the difference is your vantage point.

Run the same tests with your AI agent

Your own account, the same tests, compared with ours

  1. 1

    Get the framework

    The same open-source code that runs Atlas.

    git clone https://github.com/shifter-io/atlas
    View on GitHub
  2. 2

    Open it in your AI coding agent

    Any agent that can read a repository and run commands will do.

    • Claude Code
    • Codex
    • Cursor
    • GitHub Copilot
    • Windsurf
  3. 3

    Paste this prompt

    It sets up your account, runs the tests and writes the report. Your credentials stay on your machine.

    prompt
    This repository is the open-source test framework behind Atlas, Shifter's proxy index. Help me run the same tests on my own proxy account.
    
    1. Read the README and the methodology docs, then tell me the plan in a few lines.
    2. Ask me for my proxy provider, gateway host, port and credentials. Keep them only in .secrets/credentials.toml (from credentials.toml.example), never committed, printed or logged.
    3. Run each target of `atlas targets` once with the provider's config, as the README shows.
    4. When it finishes, give me a short report: the share of requests that passed per target, latency p50 / p90 / p99 and unique IPs, next to that provider's latency and unique IPs from https://atlas.shifter.io/v1/dataset, and its success scores as Atlas ranks them.
    5. Explain any difference in plain language.
  4. 4

    Compare with ours

    If the shapes match and only the absolute numbers differ, the method is sound and the gap is your location.

    Download the dataset
Get started

Run the method on your own account

Every rule on this page is public. Order with a 3-day money-back guarantee and test Shifter with the same targets, checks and scoring.

3-day money-back guarantee. Cancel anytime.