A residential proxy provider does not own the home internet connections it sells access to; it obtains that access from device owners, from app developers who bundle a proxy SDK, from ISPs, or from another provider that did one of those things. Datacenter proxies come from leased server ranges; mobile proxies from racks of SIM cards and modems. Understanding this supply chain answers three questions buyers rarely think to ask: why dozens of brands sell suspiciously similar pools, why “pool size” is a weaker metric than “pool freshness,” and how to tell an ethically sourced network from one built on unwitting users. It also explains a good deal about price and quality.

Datacenter IPs: leased ranges

The simplest supply chain. Hosting companies and cloud providers own large IPv4 blocks; proxy providers lease servers or address ranges from them (or hold their own ranges, announced through their own ASN), install proxy software, and sell access. Cost is server rental plus IP scarcity — IPv4 addresses are a finite, traded commodity, and clean ranges command a premium over ranges with abuse history. Because the provider controls the machines, datacenter proxies are stable, fast and cheap; because the ranges are registered to hosting entities, they are the easiest for websites to classify. The comparison with other types is in Residential vs datacenter vs mobile.

Residential IPs: four ways in

1. SDK-in-app networks

The dominant source. A proxy company offers app developers a software development kit and pays them per active device; the developer embeds it in a free app or game; when the user accepts the app's terms, their device becomes an exit node, and part of their bandwidth is resold as “residential proxy traffic.” Done properly, the user is told plainly, can opt out, and gets something in return (a premium feature, an ad-free tier). Done badly, the disclosure is a line in a 6,000-word terms document and the user has no idea their connection is being sold. The largest residential pools — tens of millions of IPs — are built this way, and the ethics vary enormously between operators using the identical technical model.

2. Paid opt-in (bandwidth-sharing) networks

Users install a dedicated app to sell their idle bandwidth for money. Consent is explicit and the economics are visible to the user. Pools are smaller and skew towards countries where the payout is meaningful relative to income, which shapes geographic coverage.

3. ISP partnerships and leases

Providers lease address blocks that remain registered to a consumer ISP and announce them from their own infrastructure. This yields ISP proxies (static residential): residential classification with datacenter reliability, and no third-party device involved. Supply is limited by which ISPs will deal and how much space they have; see What are ISP proxies.

4. Resale

Many “providers” own no network at all. They buy wholesale access to another company's pool, add a dashboard and a brand, and resell. This is legitimate and common — it is how the market reaches many buyers — but it means that when you compare five residential providers, you may be comparing five front ends to two or three underlying pools. Same exits, same abuse history, same freshness problems; different logos.

The one you must avoid

Botnets: devices compromised by malware, sold as residential exits without any consent at all. Historically part of the market's underside, and periodically dismantled by law enforcement. A provider that cannot describe its sourcing at all is a red flag for exactly this reason. Buying from such a network is not just unethical; it exposes you to IPs already flagged by every threat-intelligence feed. The ethical dimension is discussed further in Are proxies legal.

Mobile IPs: SIM farms and modem racks

Mobile proxies come from physical hardware: racks of USB modems or dedicated 4G/5G routers, each with a SIM card, hosted in a location with cellular coverage. The provider owns (or rents) the equipment and the SIM subscriptions and sells one modem's connection as a “port.” Rotation is done by forcing the modem to reconnect and receive a new IP from the carrier's pool. Because carrier-grade NAT puts thousands of real subscribers behind each public IP, these addresses have the highest trust of any type — and the smallest inventory, because every port is a physical device with a monthly SIM bill. There is also an SDK-on-phones variant, which is really the residential model on cellular connections and inherits its consent questions.

Why the supply chain shows up in your results

  • Pool size vs freshness. A 50-million-IP pool sounds vast; but the same pool is being used, right now, by every reseller's customers against the same popular targets. What matters is whether the IPs you get have been hammered on your target recently. Ask providers about per-target usage limits and how they rotate stock, not about the headline number.
  • Session stability. SDK-sourced IPs disappear when the phone sleeps or the laptop closes; that is inherent to the model, not a provider bug. ISP proxies exist to solve exactly this.
  • Geographic skew. Where the SDK apps or payout networks are popular determines where the IPs are. Coverage maps reflect the sourcing, not the demand.
  • Price floors. Per-GB pricing on residential exists because the provider is paying (in cash or in-kind) for other people's bandwidth per byte. Very cheap residential traffic is a signal to ask harder questions about sourcing; see the pricing guide.
  • Reputation contagion. Because a few wholesale pools serve many brands, a burst of abuse through one reseller degrades the IPs for everyone downstream. Provider-level abuse controls protect you as much as them.

What to ask a residential or mobile provider

  1. How do you source IPs? Expect a specific answer: SDK partners, opt-in app, ISP leases, or a named upstream. “Ethically sourced” without detail is marketing.
  2. How do device owners consent, and can they opt out? A real programme has a public page for participants.
  3. Do you own the network or resell? Both are fine; you want to know which, so you can judge whether two shortlisted providers are the same pool.
  4. What abuse controls exist? Blocking of certain targets, per-customer rate limits, KYC on customers — these keep the pool clean for you.
  5. How do you handle exit-IP freshness on high-demand targets? Any thoughtful answer beats none.
  6. For mobile: real SIMs in your own hardware, or an SDK on users' phones? Different products, different trust and stability.

Then verify what you can yourself — WHOIS the sample IPs, check classification and blocklists (how), and measure over a week (how). Our reliability ranking weighs sourcing transparency alongside the measurements.

The market in one picture

TypePhysical sourceWho controls the exitConsent question?Inventory
DatacenterLeased servers / owned rangesProviderNoVery large
ISP (static residential)ISP-registered ranges on provider serversProviderNoLimited
Rotating residentialUsers' devices via SDK / opt-in appDevice owner (indirectly)Yes — centralVery large, volatile
Mobile (hardware)Modem/SIM racksProviderNoSmall
Mobile (SDK)Users' phones via SDKDevice ownerYesMedium, volatile
Reseller (any type)Someone else's poolUpstreamInheritedAs upstream

Frequently asked questions

Where do residential proxy IPs come from?

Mostly from consumer devices whose owners — knowingly or not — run an SDK bundled in an app or a bandwidth-sharing program; also from ISP-leased ranges (ISP proxies) and from resale of other providers' pools.

Are residential proxies made from botnets?

Reputable ones are not; they use consented SDK or opt-in networks or ISP leases. Botnet-sourced pools have existed and are periodically shut down. A provider unable to explain its sourcing is the warning sign.

Why do different proxy providers give me the same IPs?

Because many brands resell the same wholesale pool. Ask whether a provider owns its network; if two shortlisted vendors are resellers of the same upstream, you are not really comparing alternatives.

Do device owners get paid?

In opt-in networks, yes, directly. In SDK-in-app models the app developer is paid and the user gets the app or a feature; whether that is fair compensation depends on how clearly it is disclosed.

How are mobile proxies made?

Primarily from racks of modems with real SIM cards operated by the provider; each modem is sold as a port and rotates IP by reconnecting to the carrier. Some are SDK-on-phone networks instead.

Does sourcing affect proxy quality?

Directly. It determines session stability, geographic coverage, how many customers share the same exits, and how quickly IPs get burned on popular targets — and it is the whole of the ethics question.