Promo code CAPROXY10 gives 10% off one purchase (orders from $10).

Proxmint

(2026)
✓ Tested by us · November 2026
  • 7-day refund window
  • Up to 31% cheaper per GB at volume
  • Low price per gigabyte
  • Over 90 million IPs
  • Traffic packs never expire

  • No refunds at all
  • Cryptocurrency payment only
  • No static or per-IP products
  • Sub-country targeting doubles traffic use
Proxmint in short. Rotation Datacenter Proxies from $0.63/GB (prices checked October 6, 2026). Our test (November 2026), Rotation Residential Proxies: 99% successful requests, 1252 ms response, 6.6% of IPs on blocklists.
Promo code CAPROXY10 gives 10% off one purchase (orders from $10).

Proxmint sells rotating proxies only, and every one of them is billed by traffic: residential opens at $1.80/GB and the grid bottoms out at $1.25/GB, while the rotating datacenter pool starts at $0.97/GB. Prices verified in October 2026.

The provider launched in 2026 out of Indonesia and compensates for the short track record the way new names do: a large claimed pool of 90M+ addresses and aggressive pricing on all three rotating products. Paid traffic never expires, so a pack bought today can be burned at whatever pace the work needs. What you give up is comfort. Crypto is the only payment rail, there is no cost-free trial and no refund window, and nothing in the catalog is static, so accounts and long-lived sessions are not what this service is built for.

Our Test

Measured November 20, 2026 · 200 requests per proxy type from our server in London
Proxy type
Success
Response, msnew / open
Speed, Mbit/s
Blocklists
Zillow
Reddit
Rotation Residential Proxies
99%
1252 / 202
3.8
6.6%
100%
87.5%
Show detailsShown below
All providers we have tested, side by side: Proxy benchmark.
Details: Rotation Residential Proxies
Reliability & speed
99%
successful requests198 of 200 requests got a response
1252 ms
response on a new connectionevery request opens a new connection — your number if you need a fresh IP each time; half answered faster; the slowest 10% took over 2933 ms. #8 of 10 tested providers of this type
202 ms
response on an open connectionrepeat requests over a connection that is already open — your number for sessions and static IPs. The IP stayed the same within a connection in 100% of requests.
3.8 Mbit/s
download speedmedian of 10 downloads of 2 MB; connection setup not counted
Response time distribution, ms
<300
43
300–600
33
600–1k
70
1–2k
50
2–5k
2
>5k
Failed requests: HTTP 403 1 · HTTP 503 1
History
October 11, 202699.5%998 ms
October 19, 202698.5%1070 ms
October 27, 202699.5%1247 ms
November 4, 202699%1300 ms
November 12, 202698%1416 ms
November 20, 202699%1252 ms
IP quality
elite
anonymityadds no headers that reveal a proxy or your IP
6.6%
IPs on abuse blocklists13 of 198 exit IPs are in FireHOL public blocklists
1%
IPs from datacenter networks2 of 196 exit IPs are in hosting or cloud networks
100%
unique IPs198 different IPs in 198 answers. 196 different /24 subnets · 124 networks
Real websites
Zillow8 / 8
8 of 8 visits opened normally
Reddit7 / 8
7 of 8 visits opened normally · blocked 1
Where the IPs come from
Exit countries we saw
United States 30Vietnam 16Netherlands 13Mexico 8Philippines 7Bangladesh 7Spain 7Saudi Arabia 6+50 more
Networks the IPs belong to
RCN17
Autonomous System number for Cyber Space,13
VNPT9
Orange5
Philippine Long Distance Telephone4
FPT Telecom4
+116 more
We paid for this access ourselves.Raw test data (CSV)Test code on GitHub
How this test works
Each request opens a new connection through the proxy and goes to ipinfo.io, which reports the exit IP, its country and its network. If that IP has used up ipinfo's limit, the request goes to ipify.org, which reports only the IP. Rotating proxies get 200 requests, static ones 50 — every static IP is also checked once a day.
A request counts as successful if an answer came back within 20 seconds.
Response time is measured from our server in London: compare it between providers, not with other people’s tests.
The datacenter share is a lower bound: we match network names against a list of hosting and cloud companies.
We repeat the test every week while we have access. Results older than 12 months are removed.
In the raw data the last part of every IP is hidden: residential proxies exit through real people’s connections.
Zillow and Reddit are checked with plain requests: both decide by the IP's reputation — Zillow blocks our own datacenter server but lets clean residential IPs through — so the result is about the proxy, not our client. A block page counts as blocked; a page that did not load is not counted. If a site blocks every request of a run, it is scored only if it let proxies in during the last 48 hours. Amazon is not tested: it showed a CAPTCHA to almost every proxy IP and did not tell providers apart.
Anonymity is checked over plain HTTP: we look for headers such as Via or X-Forwarded-For and for our own IP.
Speed: ten 2 MB downloads from our site through the proxy, counting only the time the data flows, not the connection setup. Blocklists: FireHOL level 1–3, public lists of IPs seen in attacks and abuse.
Open connection: 3 connections with 10 requests each; the first request of every connection (the one that connects) is not counted.

The code of this test is open — run it on your own proxies: github.com/caproxy-com/proxy-benchmark

How we rate providers

Pricing

Prices checked against the Proxmint website:

  • Rotation Residential Proxies
  • Rotation Mobile Proxies
  • Rotation Datacenter Proxies
1 GB
$1.80/GB/Month
5-8 GB
$1.75/GB/Month
10-13 GB
$1.69/GB/Month
25-38 GB
$1.62/GB/Month
40-60 GB
$1.53/GB/Month
100-150 GB
$1.44/GB/Month
250-350 GB
$1.33/GB/Month
500-600 GB
$1.25/GB/Month
1000 GB
$1.25/GB/Month
1 GB
$4.50/GB/Month
5-8 GB
$4.37/GB/Month
10-13 GB
$4.23/GB/Month
25-38 GB
$4.05/GB/Month
40-60 GB
$3.83/GB/Month
100-150 GB
$3.60/GB/Month
250-350 GB
$3.33/GB/Month
500-600 GB
$3.06/GB/Month
1000 GB
$2.70/GB/Month
5-8 GB
$0.97/GB/Month
10-13 GB
$0.94/GB/Month
25-38 GB
$0.90/GB/Month
40-60 GB
$0.85/GB/Month
100-150 GB
$0.80/GB/Month
250-350 GB
$0.74/GB/Month
500-600 GB
$0.68/GB/Month
1000 GB
$0.63/GB/Month

Specifications

Date established
2026
Headquarters
Indonesia
IP address rotation
yes
Money-back period (hours)
168
Free trial
no
Pool of IPs
90M+
Number of countries
195+
Targeting by city
yes
API
yes

Proxmint in our rankings

Proxy Type

  • Mobile Proxies
  • Residential Proxies
  • Datacenter Proxies
  • Rotation Residential Proxies
  • Rotation Mobile Proxies
  • Rotation Datacenter Proxies

Available Countries

Interface Language

  • English

Payment Methods

  • Crypto

Other Features

  • Rotating residential, mobile, datacenter
  • Traffic packs never expire
  • HTTP, HTTPS and SOCKS5
  • Per-request rotation or sticky sessions
  • State, city and ASN targeting
  • Login and password authorization
  • Payment in cryptocurrency

Top 9 alternatives for Proxmint

Our Review

Last updated: 2026-11-20
Published: 2026-10-06
1013

Who Proxmint fits and who should skip it

Proxmint has one billing model: you pay for traffic, never for an address. The entire catalog rotates, and that single fact tells you the audience. This is a tool for collecting data at volume and for checking what prices and ads look like from somewhere else.

The split inside the catalog is the part worth thinking about. Residential traffic is what you aim at targets that fight back, because the exit address belongs to a home connection and the anti-bot has to decide whether to punish a real customer. The rotating datacenter pool is what you aim at everything else, and everything else covers a larger share of real scraping work than most buyers admit out loud.

Skip Proxmint if you need one trusted address that stays yours for weeks. Account farming, marketplace logins, sneaker and ticket drops all want an IP that does not rotate underneath you, and there is nothing static here: no ISP proxies, no dedicated datacenter IPs, no per-IP billing at all. Sticky sessions do not close that gap either, because Proxmint caps holding one exit IP at two hours. Skip it also if your team pays for infrastructure with a corporate card.

What it actually costs you

Residential traffic starts at $1.80/GB for a single gigabyte. The number that decides your order size is where discounting stops: the last real cut lands at the 500-600 GB band, and ordering 1000 GB buys no further reduction per gigabyte.

Turn that into a month you can picture. At 120 GB of residential traffic you pay about $173, and that is the figure to carry into every comparison below. Per-gigabyte prices read easily and mislead just as easily.

The shape of the grid is what you buy on, and the shape here is a steep early slope with a flat tail.

1 GB 1.8$/GB, 5-8 GB 1.75$/GB, 10-13 GB 1.69$/GB, 25-38 GB 1.62$/GB, 40-60 GB 1.53$/GB, 100-150 GB 1.44$/GB, 250-350 GB 1.33$/GB, 500-600 GB 1.25$/GB, 1000 GB 1.25$/GB0121.8$/GB1 GB1.69$/GB10-13 GB1.62$/GB1.53$/GB40-60 GB1.44$/GB1.33$/GB250-350 GB1.25$/GB1000 GB

Inside Proxmint itself the spread between products matters more than any discount. Mobile traffic opens at $4.50/GB and the rotating datacenter pool at $0.97/GB, so picking the wrong pool costs you more than picking the wrong provider. Mobile earns its premium only against sites that treat carrier addresses gently, because hundreds of real subscribers sit behind the same IP and a ban hurts the site too. Against a target that checks nothing, the datacenter gigabyte does identical work.

There is one surcharge the price list does not show. Country targeting is included, but narrowing to a state, city, ZIP code or ASN meters traffic at double rate, so one delivered gigabyte eats two from your pack. In practice that turns the $1.44/GB you pay in the 100-150 GB band into an effective $2.88/GB the moment your job needs a city. If city-level work is your normal mode, budget for the double rate from the start, because at that point Proxmint stops being the cheap option.

Where Proxmint loses to the named competitors

The providers below sell the same product on the same unit, a rotating residential gigabyte, so the comparison holds at identical volume bands. These are not the alternatives listed higher on the page at random: each one has a traffic-billed residential pool priced in the same range.

Rotation Residential Proxies, $/GB 40-60 GB 100-150 GB 1000 GB
Proxmint $1.53 $1.44 $1.25
NodeMaven $3.98 $3.75 $2.75
Proxy Seller $2.30 $2.20 $1.45
Proxyscrape $4.25 $3.50 $2.00
Thordata $1.50 $1.00 $0.73
Geonix $2.10 $1.80 $0.75

At small and medium volume Proxmint sits near the floor of this group, close to Thordata and far under NodeMaven and Proxyscrape. Scale up and it reverses: Thordata and Geonix keep cutting into the hundreds of gigabytes while Proxmint has already gone flat, and at a terabyte a month both undercut it by a wide margin. That flat tail is the practical ceiling of this provider.

The grouped bars show where the advantage flips, and it flips somewhere between a medium month and a terabyte.

40-60 GB Thordata 1.5$/GB, 40-60 GB Geonix 2.1$/GB, 40-60 GB Proxmint 1.53$/GB, 40-60 GB Proxy Seller 2.3$/GB, 40-60 GB Proxyscrape 2.65$/GB, 40-60 GB NodeMaven 3.18$/GB, 100-150 GB Thordata 1$/GB, 100-150 GB Geonix 1.8$/GB, 100-150 GB Proxmint 1.44$/GB, 100-150 GB Proxy Seller 2.2$/GB, 100-150 GB Proxyscrape 2.2$/GB, 100-150 GB NodeMaven 3$/GB…0241.52.11.532.32.653.1840-60 GB11.81.442.22.23100-150 GB0.730.751.251.451.52.21000 GBThordataGeonixProxmintProxy SellerProxyscrapeNodeMaven
Rotation Mobile Proxies, $/GB 40-60 GB 100-150 GB 1000 GB
Proxmint $3.83 $3.60 $2.70
NodeMaven $3.98 $3.75 $2.75
Proxyscrape $4.25 $3.50 $2.00

On mobile traffic these three are close enough that price should not decide it: Proxmint leads at small volume by cents and loses at large volume. Proxy Seller and Geonix are missing from the table on purpose, because they rent mobile proxies per port per month, and a monthly port is a different product with a different risk profile.

The rotating datacenter pool has no counterpart in this group at all. Everyone else here who sells datacenter addresses rents them per IP per month, so there is no honest column to put $0.80/GB into. Compare them by task instead: rented IPs give you a fixed, known set of addresses you can whitelist and reuse, a traffic pool gives you churn, no commitment and a bill that follows your actual usage.

Here is where we disagree with the standard advice. "Buy residential for everything, it is safer" is a line that sells expensive traffic, and on most jobs it is simply wrong: a target that never inspects the exit ASN cannot tell a hosting address from a home one, so the residential premium buys nothing. The caveat is just as mechanical. The moment the target does inspect, your block rate climbs and retries eat the saving, which is exactly why you test one gigabyte against your real target before you buy a hundred.

What comes with the gigabyte, and what a test will cost you

A cheap gigabyte usually means a thin pool, so that is the first thing to check on a 2026 launch. Proxmint claims 90M+ addresses, which puts it above Proxy Seller, NodeMaven and Geonix and below Proxyscrape and Thordata. Read it as reassurance, not proof: a headline pool says nothing about how many addresses are live in the country you need, at the concurrency you run, on the day you run it.

The cost of finding that out is where Proxmint is genuinely weak. There is no cost-free trial and no refund window, and the closest thing to a trial is a one-time bonus gigabyte of residential traffic added after a first wallet top-up of $10 or more. Your first purchase is the test. The provider's terms leave no ambiguity about what happens if the pool disappoints: "Cryptocurrency payments are irreversible and, once confirmed on-chain, can never be refunded".

Provider Pool Free trial Refund window Payment methods
Proxmint 90M+ No No Crypto
NodeMaven 30M+ $3.50 No Crypto, Visa/Mastercard
Proxy Seller 47M+ No 24 h Crypto, Visa/Mastercard, PayPal, Volet
Proxyscrape 120M+ Yes 72 h Crypto, AliPay, Visa/Mastercard, PayPal, Doku
Thordata 100M+ Yes Available Crypto, AliPay, Visa/Mastercard
Geonix 15M+ $1.99 24 h Crypto, Visa/Mastercard

Read the two middle columns together, not separately. A paid trial with a refund window behind it risks less of your money than a free trial with no way back, and Proxmint offers neither. The workaround is mechanical: buy the smallest band, burn it against your real target within a day, and only then size a monthly order. Packs do not expire, so a test band that runs longer than planned costs you nothing extra.

Payment is the other hard edge, and it comes with a contradiction worth knowing before you plan a budget. The pricing page advertises card payments, while the provider's own data processing agreement states that "We do not currently accept card payments, so no card processor or merchant of record is engaged". Treat crypto as the only rail: fine for solo operators and arbitrage teams that already hold stablecoins, a blocker for anyone who needs an invoice. Our roundup of providers that accept crypto shows how rare it is to accept nothing else.

What is behind a 2026 launch

Proxmint does not run its own network. Its terms of service are frank about it: "Because the network is provided by an upstream supplier, your use is also subject to that supplier’s acceptable-use rules". In practice Proxmint is a storefront and billing layer over a larger network, which explains both a 90M+ pool on a 2026 launch and a price near the floor of this group.

It also means part of your limits is set by a company you never signed with. Banks, payment services and every government domain are blocked by default, and most destination ports are closed, SOCKS5 included. If your job hits one narrow target, check the blocked list on Proxmint's support page before you top up.

What users say about Proxmint

On Trustpilot the praise is about the two things a cheap traffic seller is bought for. One user calls the proxies extremely fast and reliable, another says they needed a cheap proxy that works over both SOCKS5 and HTTP and found it here.

The complaint worth reading is about the cheapest pool. One buyer writes that datacenter latency was so high the proxies were "completely unusable" for them, and the company answered by offering to convert the purchase to residential traffic. That fits the advice above: test the datacenter gigabyte against your real target before you size an order.

Read that page with one caveat that Trustpilot prints itself: the company has not invited customers to review it recently, so the reviews there may not be representative. For how the paid pools behave under load, see the Our Test block higher on this page.

Two scenarios with numbers

Scenario A, bulk collection from targets that do not inspect anything: internal dashboards, documentation, open catalogs, lenient search results. Take the rotating datacenter pool in the 100-150 GB band at $0.80/GB, and 120 GB of pages costs $96 for the month. Nothing in that job rewards a home address, and paying residential rates for it would almost double the bill.

Scenario B, checking ad placements and prices from several countries on a target that does fight back. Residential traffic in the same band runs $1.44/GB, so 120 GB lands at about $173 a month. Keep targeting at country level if you possibly can. Pin the same job to cities and the pack drains twice as fast, which puts the month near $346 for the same delivered volume.

What the terms say and how you connect

Everything you buy here is treated as a prepaid digital good that is provisioned immediately, which is why the refund answer is a flat no rather than a window. Under the service terms there are exactly two grounds on which a refund is even considered: a verified duplicate or technically erroneous charge, and a verified total failure to deliver the service caused solely by the provider and left unresolved. Neither of them covers "the pool did not work well enough for my target".

Do not reach for a card chargeback as a shortcut either. The terms say that in that case "we may immediately suspend your account, reverse and reclaim the provisioned bandwidth, and remove the corresponding wallet credit", which is published in the terms of service. Support first, disputes never.

Two more clauses change who can buy. The marketing pages promise that an email address and a payment are all the service ever asks for, while the terms reserve the right to demand identity or business verification where use raises compliance concerns and to suspend access until it is completed. Russia, Belarus, Iran, North Korea, Cuba, Syria and the restricted regions of Ukraine are off limits as both origin and target, and any dispute goes under Indonesian law, which is a long way from a small European claim.

Verdict

Worth buying if you run a steady medium volume of rotating traffic, crypto is not a problem, and you need no static addresses. On tens to a few hundred gigabytes a month Proxmint is one of the cheapest ways to get both a residential pool and a per-gigabyte datacenter pool from the same dashboard, and the fact that packs never expire removes the usual penalty for buying a bigger band.

Not worth buying if you pay through a company account, if you want any kind of refund safety net, if you consume a terabyte a month, or if your work needs city and ASN precision as a default rather than an exception. Each of those turns the cheap gigabyte into either an unusable one or a doubled one, and other rotating residential providers will serve you better.

Sources

Promo code CAPROXY10 gives 10% off one purchase (orders from $10).
Visit Website Proxmint

FAQ

  • What types of proxies does Proxmint sell?
    • Three rotating products, all billed by traffic: residential from $1.80/GB, mobile from $4.50/GB and a rotating datacenter pool from $0.97/GB. There are no static products, no ISP addresses and no per-IP billing.
  • Is Proxmint suitable for multi-accounting?
    • Not really. A sticky session holds one exit IP for up to two hours, and accounts usually need an address that stays the same for weeks. For that you want ISP or dedicated proxies from a provider that sells them per IP.
  • Is there a free trial?
    • No. The advertised free gigabyte is a one-time bonus of residential traffic unlocked after a first wallet top-up of $10 or more, so your first paid purchase is the test.
  • Does Proxmint refund unused traffic?
    • No. Packs and wallet top-ups are treated as prepaid digital goods, crypto payments are irreversible once confirmed on-chain, and the terms allow a refund only for a duplicate charge or a total service failure caused by the provider.
  • What payment methods are accepted?
    • Cryptocurrency only. The pricing page mentions cards, but the provider's own data processing agreement states that card payments are not accepted at this time, so plan for crypto.
  • Can I target a specific city or ASN?
    • Yes, on all three pools, but that traffic meters at double rate: one delivered gigabyte uses two from your pack. Country targeting is included at the normal rate.
  • Does the purchased traffic expire?
    • No. Packs carry over, so buying a larger band for the lower per-gigabyte rate does not risk losing the unused part.
  • How good are the proxies in practice?
    • Judge them by measurement rather than marketing. The Our Test block higher on this page shows success rate, response time, rotation behavior and reputation checks for Proxmint's pools, and it is refreshed regularly.

Reviews (0)

Leave a review

Rating Proxmint:
Proxmint
Reviews (0) Leave a review https://proxmint.com/