What Is a Proxy Pool?

The three terms overlap. The difference is who holds the IPs and who decides which one you use.

A provider's "proxy network" usually means its pools of different types, residential, mobile and datacenter, and the gateways in front of them. At ProxyField that is 412M+ residential and 10.2M+ mobile IPs in 195+ countries and 21,000+ cities, reached through one gateway.

You send every request to the gateway with your credentials and, optionally, targeting and session settings. For each request the gateway:

1. Checks who you are, by username and password or by your allowlisted IP.
2. Filters the pool to IPs that match your target: country, region, city.
3. Picks an IP, a new one per request, or the one tied to your session for a sticky session.
4. Forwards the request through that IP to the site, and the reply comes back the same way.

Behind the scenes, a provider keeps its pool healthy: it tracks which IPs are online and how they perform, and stops sending traffic to slow or failing ones. ProxyField's gateway, gate.proxyfield.io:7000, gives you a new IP on every request or the same IP for a sticky session from 30 s to 36 h, and "if a node slows down, traffic automatically moves to the next one." Rotation and sessions are compared in sticky vs rotating proxy sessions.

"Backconnect proxy" is an older industry name for the gateway in front of a pool. Your app connects to one fixed address, and the gateway connects "back" to whichever exit IP in the pool will carry the request, so the address you configure never changes while the exit IPs do. Rotating, backconnect and gateway proxies usually describe the same setup from different angles: rotating proxies covers the product side.

A backconnect gateway can sit in front of any IP type. Residential and mobile pools are almost always reached this way, because their IPs come and go; datacenter pools can be too. The extra hop through the gateway adds a little latency, usually far less than the time a blocked request costs.

Use a pool when you want many IPs without managing them, and a list when you want to control a fixed set of IPs yourself.

• A pool suits scraping, rank tracking and price checks: one endpoint, a new IP per request, any country or city, and you pay for traffic. The provider replaces IPs that go offline.
• A list suits tools that need a fixed set of IPs, such as allowlisted dashboards or one dedicated IP per account. You decide which IP each task uses, and you build the rotation and health checks yourself.

Some tools ask for a list even when you use a pool. ProxyField's configurator generates as many proxy lines as you need, and each line points at the gateway, so a tool that wants 100 proxies gets 100 lines with the pool behind them. Dedicated IPs for a list are covered in shared vs private vs dedicated proxies.

Yes, but the headline number matters less than how many IPs exist where you need them. A pool of millions helps only if it has enough IPs in your target country and city to keep each one lightly used.

What to check:

• IPs per country and city. A US campaign needs US depth: ProxyField lists 61.2M+ residential IPs in the United States, 16.6M+ in the United Kingdom and 20.2M+ in Germany. Per-country figures are on each location page.
• Your concurrency against that depth. 500 parallel threads on a city with a few hundred available IPs will reuse addresses quickly. Narrow targets need slower rates or wider targeting.
• Repeat IPs in a test. Send a few thousand requests to an IP-echo endpoint and count unique IPs; many repeats mean a shallow pool for that location.
• Success on your target. A smaller pool of clean IPs beats a larger one with a poor reputation, which only a test on the real site shows.

If you own a list of dedicated proxies, you can turn it into a small pool in code: pick an IP per request, skip IPs that failed recently, and put them back after a cool-down.

import random, time
import requests

PROXIES = ["http://<login>:<password>@203.0.113.10:8000", "http://<login>:<password>@203.0.113.11:8000"]
resting = {}  # proxy -> time it may be used again

def fetch(url):
    ready = [p for p in PROXIES if resting.get(p, 0) <= time.time()]
    proxy = random.choice(ready or PROXIES)
    try:
        r = requests.get(url, proxies={"http": proxy, "https": proxy}, timeout=30)
        if r.status_code in (403, 429):
            resting[proxy] = time.time() + 300  # rest a blocked IP for 5 minutes
        return r
    except requests.RequestException:
        resting[proxy] = time.time() + 60
        raise

This is the work a provider's gateway does for you at scale, with far more IPs. More on running your own setup in the build your own proxy guide, and on managing many proxies in proxy management.

Need a pool instead of a list? ProxyField residential proxies: 412M+ IPs in 195+ countries and 21,000+ cities behind one gateway, with a new IP per request or sticky sessions up to 36 h, from $1.95/GB with no subscription. Start with the Easy Trial: 1 GB for $2.95. Traffic never expires, and all purchases are final.