How proxy pricing works

Proxy quotes are hard to compare because the unit of billing changes. Some plans charge per IP per month, others per gigabyte, and others bundle volume, regions, and protocols. Understanding the billing model matters more than chasing the lowest headline number.

Three common ways prices are quoted

Per IP or per port, usually billed monthly

Datacenter proxies and many ISP / static residential products use this model: you rent a number of relatively stable IPs for a term, often with a concurrency or bandwidth cap. It fits work that needs a steady exit and is limited more by connections than by page weight. When you compare plans, check whether IPs are exclusive or shared, whether regions are fixed, how overages work, and whether unused time expires.

Per gigabyte

This is the usual sticker price for rotating residential and mobile proxies. The unit price often falls as the plan gets larger, while on-demand rates for small volumes can be much higher. Ask whether there is a minimum purchase, whether unused traffic expires, and whether failed requests or response headers are billed. A low unit price can still produce a high monthly bill.

Bundled or platform fees

Some providers fold traffic, concurrency, regions, and extra APIs (for example parsing or unlocking) into a monthly fee. That can be simpler, but you should strip out features you will not use before you compare. Free credits, new-customer deals, and coupon codes usually expire and should not be treated as a long-term cost.

Easy mistakes when comparing quotes

Compare the same product type. A few dollars per datacenter IP per month is not the same unit as a few dollars per GB of residential traffic.
Region and quality are often outside the sticker price. Two “US residential” plans can differ in available subnets, city-level targeting, and session stability.
Taxes and payment fees. Prices may exclude VAT, and some payment methods add their own fees.
Directory starting prices are only a lead. When this site shows a starting price, it is because a public source existed. The date and the way the price was quoted may differ. It is a clue, not a checkout price.

A steadier way to compare

Lock the product type and regions first, then collect quotes for that type only.
Convert each quote into the same unit based on your expected monthly usage (GB or number of IPs).
Add the extras you will actually need: sticky sessions, city targeting, SOCKS5, or invoicing.
Run a real test with the smallest purchase unit before you scale up.

This site does not publish its own speed tests or value rankings. Price labels in the directory are there to give you a rough order of magnitude, not to replace an official quote.