Residential, data center and mobile proxies
The three kinds differ in the network your traffic leaves from. A data center proxy exits from a server at a hosting company, a residential proxy from a home connection at a broadband provider, and a mobile proxy from a cellular carrier's network. Sites tell them apart by who owns the address and how it is used, and give each kind a different level of trust.
How each kind gets its address
Data center
A hosting company owns address ranges and announces them on the internet under its own autonomous system number. An address is assigned to a server or virtual machine and usually stays with it for as long as the contract runs. When a proxy provider buys hundreds of addresses from the same host, they often come as neighbours in one range.
Residential
A home broadband provider gives the subscriber's router a public address when it connects, for example over PPPoE or DHCP. Many providers change it when the line reconnects, and behind it there is usually one household. Some providers now put home lines behind shared address translation too.
A residential proxy means your traffic really leaves from a device on a home connection. Someone's device is relaying your requests, which raises the sourcing question covered below.
Mobile
A phone on a 4G network gets an internal address from the carrier. It may come from private ranges or from 100.64.0.0/10, the block that RFC 6598 reserves for provider networks using carrier-grade NAT (CGN). On the way out, the CGN translates it to a public address that a large number of subscribers share. The mechanism is explained in What is CGNAT.
Side by side
| Data center | Residential | Mobile | |
|---|---|---|---|
| Network owner | Hosting company | Home broadband provider | Mobile carrier |
| Who uses the address | One server | Usually one household | Many subscribers at once |
| Address stability | Usually fixed for months | Changes on reconnect with some providers | Rotated by the carrier |
| How sites see it | Caution or blocks on sensitive sites | Treated as an ordinary user | Treated as an ordinary user; blocking is costly |
| Speed | High and steady | Follows the home line | Follows signal and network load |
| Relative cost | Usually lowest | Higher | Usually higher |
How sites tell which kind an address is
- The autonomous system owner. RFC 1930 defines an autonomous system as a set of prefixes under one routing policy. The owner's name and line of business settle most cases straight away.
- Public records. The WHOIS record names the organisation the range was assigned to, and the reverse name (PTR) sometimes contains words that point to a server or a subscriber line.
- Lists of end-user ranges. On the Spamhaus PBL, providers themselves register subscriber ranges that should not send mail directly. Being listed marks an address as end-user space, not as badly behaved.
- Behaviour. When a site sees many different accounts and browsers behind one address, it concludes the address is shared behind a CGN.
- Open ports. An address that answers on service or proxy ports points to a server, not a phone or a home router.
These signals are combined with others into one decision. More in How sites detect VPNs and proxies.
What a shared address means in practice
RFC 6888 requires a CGN to use the same external address for all sessions of one subscriber by default. The site sees one stable address for you while the mapping lasts, not a different one per request.
The same standard requires support for a per-subscriber limit on external ports. A single mobile exit therefore cannot hold thousands of simultaneous connections the way a data center server can. That makes it a better fit for a person working on a device than for heavy automated collection.
Sharing has another side: reputation is shared too. If another subscriber abuses the address, everyone behind it may face extra checks for a while. That is why large sites weigh account and device signals alongside the address, not the address alone.
When the classification is wrong
Databases classify ranges, not devices, and that produces well-known errors. A small provider may offer home lines and hosting from the same autonomous system, so its subscribers are sometimes classed as servers. A range that moves from one owner to another can keep its old label for weeks or months.
Mobile networks do not serve only phones, either. Home 4G routers exit from the same carrier ranges, so a household looks like a phone. Careful sites therefore base their decision on several signals and never on the database label alone.
Why the cost differs
Data center addresses are bought in bulk and serve a lot of bandwidth cheaply. Residential addresses need real devices on home lines, so they are usually sold by data volume. Mobile addresses need SIM cards, devices and data allowances with a carrier, which keeps the cost per gigabyte higher.
Where residential proxy addresses come from
The provider does not own these addresses; they belong to subscribers of other broadband providers. They are usually obtained through software that people install and that agrees to share their connection, sometimes in exchange for a free service. In its description of the XBL, Spamhaus mentions free VPN apps that turn the user's device into a proxy for unknown parties.
Ask the provider how it gets consent from device owners. If it cannot say, you do not know who is relaying your traffic or what happens to it on that device.
How to check what kind an address is
- Connect through the proxy and open My IP. Look at the network and ASN fields.
- Read the address's reverse name from the command line:
dig +short -x 203.0.113.7A missing reverse name means nothing on its own, but a name containing a hosting company's domain usually settles it. The short checklist for a Syrian address in particular is in Syrian mobile IP vs data center IP.
Which kind to choose
- Data center for collecting public pages in bulk, testing your own sites, and any task where looking like an ordinary user does not matter.
- Residential when you need a home address in a specific country or city, provided you know where the addresses come from.
- Mobile for services built for phones, and for apps that expect a subscriber on a carrier network in a given country.
syriaproxy is a mobile proxy: its addresses come from the Syriatel 4G network inside Syria (AS48065) and are shared with ordinary subscribers behind CGN. The same address lasts about two hours before Syriatel changes it, and you can request a new one at any time.
FAQ
Is a mobile proxy faster than a data center proxy?
Usually not. Data centers sit on high-capacity links, while mobile speed follows signal quality and network load. The advantage of mobile is how it is classified, not how fast it is.
Is a data center address in Syria as good as a Syrian mobile address?
No. The country is the same, but the network owner is a hosting company, and sites read that from the ASN. The address may also be registered to Syria while the server sits in another country.
Why does only the mobile address change on its own?
Because the carrier keeps reassigning its public addresses among subscribers, as it does for every phone on the network. How long an address lasts varies from carrier to carrier.
Look up any address with the free lookup tool to see its owner, network type and ASN.