ISP vs Residential Proxies: Which One Your Job Needs

In short: A static ISP proxy is one address, held for the period you pay for and billed per address. A rotating residential proxy draws a new address per request or per short window and bills per gigabyte. Choose the static one when a target has to recognise you, and a rotating pool when each request should look like a stranger’s.

The difference between a static ISP proxy and a rotating residential proxy is whether the address stays yours, and that one property decides the price model, the failure modes and the jobs each one suits. If you want the definition first, start with what an ISP proxy is.

How do ISP and rotating residential proxies assign an address?

A static ISP proxy gives you one known address; a rotating one gives you a gateway into a pool.

With a static ISP proxy, you know the address in advance. It stays the same, nobody else uses it while you hold it, and you are billed for holding it.

A rotating residential proxy hands you nothing fixed. You connect, and the provider chooses an exit for each request: usually a different one each time, or the same one for a short window if you ask for stickiness. The bill counts the traffic you moved.

Why is a static address billed per IP and a pool per gigabyte?

The billing follows from whether an address is reserved for you. A reserved address is held out of circulation on your behalf, so a provider charges rent on it: a price per address per period that does not depend on how much traffic you push through it.

A pool reserves nothing for you, which leaves traffic as the one thing to meter. Small jobs come out inexpensive, and large ones produce a bill that depends on what the target served. A page that starts loading video halfway through a crawl turns into a budget problem you find out about afterwards.

Which jobs break when the address rotates?

Anything that depends on continuity breaks on a rotating pool. A logged-in web session is normally a cookie the site set, as RFC 6265 describes, and the cookie travels with every request, as does the address the request came from. OWASP’s session guidance recommends binding a session to properties such as the client IP address and treating a change mid-session as a sign of hijacking.

A session that begins on one address and continues from another can therefore look, from the far side, like an account handed to someone else. These fail first:

• login flows;
• multi-step forms;
• anything else holding server-side state between requests;
• any connection open long enough to outlive a sticky window.

In each case the site sees a session’s address change halfway through, the pattern OWASP tells it to treat as hijacking.

Where does a fixed address run into limits?

A fixed address struggles with breadth, because one address making many requests is one very busy visitor. Counting requests per address is a standard option in the tools sites use to limit traffic. Cloudflare’s rate limiting rules list the client IP among the characteristics a counter can be keyed on, and NGINX’s request-limiting module uses the client address as the key in its documented example.

Against a target that counts this way, a fixed address concentrates all of the attention in one place. Renting an address by the month to use it for an hour is also the wrong shape of purchase, since the rent runs whether the work does or not.

Does the target need to recognise you?

When the target needs to recognise you, you want the address to hold still. Work that gains from arriving as an unremarkable stranger each time belongs on a pool, where a fixed address would only add cost.

Most arguments about this are two people doing different jobs under one word. Scraping covers both shapes, which is why the same question gets opposite answers from people who are each right about their own work.

Some programmes need both: broad discovery to find what is worth watching, then continuous observation of the few things that were. Run the two through separate addresses. A rate limit is often counted per IP, so keeping them apart stops exploratory traffic from using up the allowance an account you run under the platform’s permission depends on.

What does ProxyJam sell?

ProxyJam sells the static side only: one address, one named carrier, held for the period. Traffic is unmetered. Heavy use affects how many connections you may hold open: past 50 GB in a calendar month, a proxy’s ceiling falls from 100 concurrent sessions to 10 until its next billing period (Terms of Service, section 2.4).

The price per IP is on the ISP proxies page and the pricing page. The same addresses appear under their other market name on static residential proxies.

We do not sell rotation yet. Its card on our site says coming soon, with no plan and no checkout behind it. If a pool is what your job needs, buy one from a provider who has it; our rotating residential page explains how the pool model works.

ISP or rotating residential: which does your job need?

Hold an address and accept a fixed bill when the target has to recognise you, and draw from a pool when each request should look new. Two other comparisons cover the neighbouring questions: residential against datacenter looks at the registration itself, and mobile against residential looks at what a carrier address adds.