Contribute to Atlas
One measurement is an anecdote.
Many that agree make a record. List your network, share your own test results, or review a provider you pay for.
- Success rate 98.9% 99.1%
- Latency p50 412 ms 391 ms
- Unique IPs 4,812 4,967
- Right country 99.6% 99.9%
Within 2 points on every figure: the shape holds
- Free to join
- Stays private
- Score unaffected
- Not auto-published
Ranked by Atlas Score across 18 networks
Shifter94.1#1
Oxylabs83.4#2
Decodo81.5#3
FleetProxy80.6#4
Geonode80.3#5
DataImpulse80.1#6
ProxyEmpire79.9#7
Webshare79.3#8
NodeMaven77.7#9
SOAX77.5#10
Proxy-Seller77.2#11
Byteful75.6#12
Infatica74.6#13
Massive74.2#14
Evomi73.1#15
Rayobyte71.5#16
Proxy-Cheap70.5#17
IPRoyal66.4#18
Shifter94.1#1
Oxylabs83.4#2
Decodo81.5#3
FleetProxy80.6#4
Geonode80.3#5
DataImpulse80.1#6
ProxyEmpire79.9#7
Webshare79.3#8
NodeMaven77.7#9
SOAX77.5#10
Proxy-Seller77.2#11
Byteful75.6#12
Infatica74.6#13
Massive74.2#14
Evomi73.1#15
Rayobyte71.5#16
Proxy-Cheap70.5#17
IPRoyal66.4#18Three ways to contribute
Pick the one that fits who you are.
-
List your network
01For proxy providers
Put your product under the same tests, targets and formula as every network in Atlas.
-
Share test results
02For anyone who tests proxies
Send your own measurements. We compare them with ours before anything is shown.
-
Review a provider
03For paying customers
Support, billing and documentation, the things no measurement can reach.
A question, a correction, or a network that wants to reply to its results? Contact the team
Next · 02 List your network How does a network get listed?Run a network? Put it under test
Your product goes through the same tests, targets and formula as every other network, Shifter included.
What we need from you
-
One real account
On the product you sell, set up the way a customer gets it.
-
A named plan
The tier is shown next to every figure on your page.
-
Kept funded
For as long as the listing runs, so results never come from a throttled account.
-
Permission to test
Written consent to automated traffic around the clock.
What happens next
- 1 You apply in the live chat or by email.
- 2 We reply there.
- 3 We set up a scoped test account over a private channel.
- 4 The first measurement window fills in 24 hours.
- 5 Your page shows what the tests show.
Read the exact rules
The rules of a listing
The account is the product as sold
No better pools, priority routing, allowlist entries or support flags. If the account is tuned, the figures describe a setup no customer can buy, which makes them worthless to everyone, including you.
A lapsed account pauses the listing
If the allowance runs out, we pause the listing and say on the page that it is paused, rather than publishing figures from a throttled account.
Each product is ranked on its own
Claiming a product you do not really sell earns a poor score in a category you never wanted to be in.
Results are published as they come
A poor result is published like a good one. You get a right of reply next to it, not a right to edit or remove it.
No credentials by chat or email
The application never asks for passwords. Access to the test account is set up over a private channel after we reply, and should be a sub-account scoped to the tested plan.
Share your test results
Our tests run from fixed locations on mid-tier plans. Yours do not, and that is why your numbers are worth having.
Step 1 of 5
Which provider did you test?
One provider per result. Only networks in Atlas can be submitted.
Read the exact rules
How we handle what you send
Files are read as data
A results file is opened as a table or as JSON and never run. Anything that is not data is set aside unread.
Rejections stay on the record
A result we cannot use is listed with the reason, so the record cannot be shaped by quietly dropping what disagrees with us.
Your identity stays private
We publish a handle of your choice, never your name, company or email.
The expected format
- A CSV or JSON export with one row per request.
- For each request: the time, the target, the HTTP status, whether it passed your check, the latency in milliseconds, and the country of the exit IP.
- One provider and one product per file. The form asks for the plan and where you tested from.
- A few thousand requests over at least a day is enough to compare shapes; a few dozen is not.
Review a provider you pay for
Support, billing and documentation are what a measurement cannot reach. Reviews never change the Atlas Score.
-
Shifter #1 · 94.1 · 0 reviews #1 94.1 A+ No reviews yet Write a review -
Oxylabs #2 · 83.4 · 0 reviews #2 83.4 A No reviews yet Write a review -
Decodo #3 · 81.5 · 0 reviews #3 81.5 A No reviews yet Write a review -
FleetProxy #4 · 80.6 · 0 reviews #4 80.6 A No reviews yet Write a review -
Geonode #5 · 80.3 · 0 reviews #5 80.3 A No reviews yet Write a review -
DataImpulse #6 · 80.1 · 0 reviews #6 80.1 A No reviews yet Write a review -
ProxyEmpire #7 · 79.9 · 0 reviews #7 79.9 B+ No reviews yet Write a review -
Webshare #8 · 79.3 · 0 reviews #8 79.3 B+ No reviews yet Write a review -
NodeMaven #9 · 77.7 · 0 reviews #9 77.7 B+ No reviews yet Write a review -
SOAX #10 · 77.5 · 0 reviews #10 77.5 B+ No reviews yet Write a review -
Proxy-Seller #11 · 77.2 · 0 reviews #11 77.2 B+ No reviews yet Write a review -
Byteful #12 · 75.6 · 0 reviews #12 75.6 B+ No reviews yet Write a review -
Infatica #13 · 74.6 · 0 reviews #13 74.6 B+ No reviews yet Write a review -
Massive #14 · 74.2 · 0 reviews #14 74.2 B+ No reviews yet Write a review -
Evomi #15 · 73.1 · 0 reviews #15 73.1 B+ No reviews yet Write a review -
Rayobyte #16 · 71.5 · 0 reviews #16 71.5 B No reviews yet Write a review -
Proxy-Cheap #17 · 70.5 · 0 reviews #17 70.5 B No reviews yet Write a review -
IPRoyal #18 · 66.4 · 0 reviews #18 66.4 B No reviews yet Write a review
- Your name and company stay private
- Providers can never alter a review
- Ratings never move the Atlas Score
- Bad reviews stay, with a reply below
Read the exact rules
What each badge on a review means
Verified use
ResultsThe reviewer also sent test results for the same provider that we could compare with ours, proof of hands-on use.
Unverified
LabelPublished, clearly labelled, and kept out of the verified average. Useful signal, weaker evidence.
Every submission, on the record
Accepted and rejected runs and reviews are listed here, with the reason for each.
No community runs published yet
Accepted and rejected results and reviews appear here once the first ones have been checked, each with its status and the reason.
Test Shifter yourself, then send us the numbers
Run your own tests and share the result here. It is checked against ours the same way as a result for any other network.