Standard residential and premium residential run on the same real-device network — which is exactly why the names confuse people. Both reach the same targeting depth: state, city, ZIP and ASN work on either one. They split on two things instead — how that targeted traffic is metered, and how tightly the pool is curated. Here is the practical difference, the cost math, and a simple rule for choosing.
The 30-second answer
Standard residential is the volume workhorse: real home ISP IPs, country targeting free, rotating or sticky sessions, billed per gigabyte. It reaches state, city, ZIP and ASN as well — that traffic simply meters at 2×, drawing 2 GB from your pack per gigabyte delivered. It is the right default for web scraping, SERP tracking, ad verification and market research at scale, and it stays the cheaper pool even when every request is finely targeted.
Premium residential is a curated, higher-reputation slice of the same network, and the one pool that meters fine geo at 1×. You are not buying access to city or ASN targeting — standard has that. You are buying exits selected for clean history and consistency, plus targeted traffic that bills flat instead of double.
What standard residential is built for
Every exit is a genuine household IP, so requests look like organic local visitors — that trust is what you are buying over datacenter. Rotation spreads volume across millions of addresses, sticky sessions hold one IP for stateful flows, and targeting is a username parameter away: cr.us for a country, or cr.us;city.newyork;zip.10001 to land in one metro. Bandwidth bills per gigabyte and never expires.
The trade-off is the meter, not the reach. Country-only traffic bills 1:1; anything carrying a state, city, ZIP or carrier bills 2:1, so a finely targeted gigabyte costs two. The pool is also tuned for breadth rather than per-request consistency — at high volume the occasional block or retry is expected and harmless. If your job tolerates that, standard is the economical choice whether or not you target below the country.
What premium actually adds
First, flat metering. State, city, ZIP and ASN targeting — alone or stacked, like cr.us;city.newyork;zip.10001 — bill at the plain base rate per gigabyte. There is no 2× burn, so a pack drains at the rate printed on it no matter how narrowly you aim. On a job that is entirely fine geo, that is the difference between a 100 GB pack delivering 100 GB and delivering 50.
Second, the pool itself is curated. Premium draws on a higher-reputation slice of the network selected for clean IP history and predictable regional behavior, so sessions are steadier and first-pass success is higher on strict targets. That is the real reason it exists as a separate tier — flat metering alone does not pay for it, as the next section shows.
Side by side
The table below is the whole decision at a glance. Same network, same gateway, same targeting reach, same pay-as-you-go billing — the tiers split on metering, consistency, and price per gigabyte.
| Residential | Premium residential | |
|---|---|---|
| IP source | Real home ISP IPs | Same network — curated, higher-reputation slice |
| Geo targeting | Country + state + city + ZIP + ASN | Country + state + city + ZIP + ASN |
| Fine-geo metering | 2× — a targeted GB draws 2 GB of balance | 1× — targeted traffic bills like any other GB |
| Sessions | Rotating or sticky (up to 120 min) | Rotating or sticky (up to 120 min) |
| Consistency | Tuned for volume; occasional retry expected | Tuned for precision and per-request reliability |
| Price per GB | Lower — the economical default | ~5× standard; ~2.5× standard's rate per delivered fine-geo GB |
| Best for | Scraping, SERP, ad verification, research at scale | Hyper-local verification, geo-pricing, compliance, brand protection |
The cost math, honestly
It is tempting to assume the 2× burn is what premium saves you from, and that a fine-geo-heavy job therefore belongs on premium. Run the numbers and it does not hold. Premium lists at roughly 5× standard's rate per gigabyte, while the surcharge only doubles standard's. Even on traffic that is 100% finely targeted — the worst case for standard — standard still delivers a gigabyte for about 40% of what premium charges, and that ratio holds at every pack size on the curve.
So premium cannot be justified on price, and we would rather say so than sell you the arithmetic backwards. What it is justified on is the pool: cleaner IP history, steadier sessions, higher first-pass success on strict targets — the hidden cost that never shows on an invoice as blocked requests, retries, and geo data collected from the wrong place. Flat billing is a convenience on top of that, not the reason to buy.
When you'll feel the difference
You will feel premium when the job is location-critical or failure-sensitive: verifying ads or prices down to a ZIP code, compliance checks that must reflect an exact locale, brand protection sweeps, and long session-dependent flows on strict targets where a mid-run block means starting over.
You will not feel it on block-tolerant volume work, targeted or not. A scraper that shrugs off a retry gets nothing from the premium tier except a bigger bill — even if every request pins a city. That workload belongs on standard, where the 2× burn is still the cheaper way to buy a targeted gigabyte.
A simple decision rule
Ask one question, not two: does a failed or misplaced request cost you more than the bandwidth it used? Yes points to premium, no points to standard. Whether you target below country level does not enter into it — both pools do that, and standard does it for less even after the 2× burn. The common shape is running the bulk on standard, fine geo included, and routing only the failure-sensitive share through premium.
And you do not have to choose once. Each pool is its own credential on the same gateway, so teams route the precision share through premium and the bulk share through standard. Bandwidth never expires — buy a small pack of each, test on your real targets, and scale whichever wins.