Home/Blog/Mobile vs Residential

Mobile Proxy vs Residential Proxy: What the ASN Changes

P
ProxyBase Team
November 2026 • 8 min read

Mobile and residential proxies are sold as two tiers of the same product, but they route traffic through different parts of the network. A residential proxy exits through an ISP address assigned to a home connection. A mobile proxy exits through a carrier gateway, usually behind carrier-grade NAT, where many phones share one public IP.

That difference sets the pool size, price, latency, and what the target sees. Which one you need depends on which failure you are paying to avoid.

The exit network defines the type

A residential proxy exits through an IP address that an ISP assigned to a real home connection. Comcast, AT&T, BT, Deutsche Telekom. The target sees a residential ASN and a normal consumer location.

A mobile proxy exits through a 4G or 5G carrier gateway. Most carriers put many devices behind one public address through CGNAT. The target sees a carrier ASN, and that address is shared with every phone behind the gateway.

Both avoid datacenter ranges, which is why either one can pass fraud checks that reject hosting IPs. The choice between them is a trade between pool size, price, and how expensive the address is to block.

Inventory is not uniform. Which seller and country bucket you land in depends on who is online and what you request. A country with deep residential supply can have thin mobile supply, and the price reflects that.

Why a carrier IP is expensive to block

A carrier ASN gets a high trust score because blocking it is expensive. A residential IP can be blocked with little collateral. A carrier NAT address sits in front of many legitimate phones, so a blanket block breaks those users too.

That asymmetry is what you are buying. The same shared address that makes a mobile IP expensive to block also means you inherit the reputation of every device behind it. A mobile IP is shared, and its reputation is not yours to control.

Residential ranges have the opposite problem. Scrapers use them heavily, and some ISPs get flagged as a group. A carrier ASN is harder to write off, which is why mobile sometimes works where residential does not.

Match the network type to the workload

WorkloadStart withWhy
Bulk scraping of public pagesResidentialLower cost per GB and a larger pool. The mobile premium rarely pays for itself.
SERP checks, price monitoring, lead enrichmentResidentialRequests are small and independent. Retries are cheap.
Mobile-first app, or a site that blocks ISP rangesMobileThe carrier ASN is the signal the target expects to see.
Login, checkout, multi-step formMobile, stickyCarrier trust plus one stable exit for the whole flow.
High-fraud-check flowsMobile, stickyFraud scoring weights carrier ASNs heavily. Keep the session stable.
A site or ad that serves carrier users differentlyMobileThe target deliberately changes the response for cellular traffic.
Geo-restriction where ISP ranges are blockedMobileCarrier range blocks are less common and costlier for the target.

If you are unsure, start residential. It is the cheaper experiment, and a residential session can still be rotated or replaced. Paying for mobile before measuring a residential block rate usually wastes money.

Count the cost per successful request

ProxyBase residential starts at about $3 per GB. Mobile starts at about $5 per GB. Rates vary by country and move with supply, so check /markets before you budget.

A 100 GB job costs $300 at $3 per GB and $500 at $5 per GB. A cheaper per-GB rate can still produce a more expensive job. Suppose residential succeeds on 80 percent of requests and retries add 25 percent bandwidth. You pay for 125 GB to deliver 100 GB of successful responses: about $375. Mobile has to cut the failure and retry rate enough to beat that, and at $5 per GB it usually does not unless the target specifically rejects residential ranges.

The usual answer is a residential baseline plus a mobile session for the subset that fails. That is two sessions against one budget.

Measure the block rate before you switch

Run the same 1,000 requests through each network type. Count 2xx, 403, 429, and CAPTCHA pages. Then compute cost per successful request: bytes consumed divided by gigabytes, multiplied by the per-GB price, divided by the success rate.

A target that returns 403 on 30 percent of residential requests is a mobile candidate. A target that returns 403 at the same rate on both types is not a proxy problem. Fix the headers, the TLS fingerprint, or the request rate first. The scraping guide covers that order of operations.

Both are session types on the same gateway

ProxyBase does not run separate endpoints for mobile and residential. The network type is a field when you open a session. Country is also set there. Neither appears in the credential string.

import requests API = "https://api.proxybase.xyz/v2" TOKEN = "sk_your_session_token" # from POST /v2/auth/verify def open_session(network_type): r = requests.post( f"{API}/sessions", headers={"Authorization": f"Bearer {TOKEN}"}, json={ "country": "US", "network_type": network_type, # "residential" or "mobile" "session_type": "sticky", "spend_cap_microcredits": 5_000_000, }, timeout=15, ) r.raise_for_status() return r.json()["session_id"] session_id = open_session("residential") proxy = f"socks5h://{session_id}:{TOKEN}@api.proxybase.xyz:1082" proxies = {"http": proxy, "https": proxy} r = requests.get("https://api.ipify.org?format=json", proxies=proxies, timeout=30) print(r.json())

Change residential to mobile and the switch is done. The gateway, port, and credential shape stay the same. The username is the session id and the password is the session token. Port 1082 is the v2 buyer gateway.

Sticky or rotating applies to both

Network type and rotation policy are independent. Sticky pins one seller for the session, which is what logins, carts, and pagination need. Rotating matches a new seller per SOCKS5 connection, which suits independent URLs.

On mobile, sticky is the common choice for account work. The carrier IP is the trust signal, and changing it mid-flow looks wrong. Rotating mobile is for distributed jobs where the carrier ASN matters but no state carries across requests.

One trap applies to both: keep-alive. A rotating session only rotates when a new connection opens. A requests.Session that reuses connections pins one exit. Send Connection: close or open a new connection when you need a fresh IP. The rotation guide has the full mechanics.

Six mistakes that waste the mobile premium

  • Buying mobile for the whole job because one host blocked residential.
  • Treating a mobile IP as private. CGNAT means the address is shared, and so is its reputation.
  • Rotating mid-login and then blaming the proxy when the session drops.
  • Comparing dollars per GB without counting retries and failed requests.
  • Ignoring latency. Mobile round trips are higher and less stable, so timeouts and concurrency need to change with the network type.
  • Expecting a proxy to fix fingerprinting, headers, or a request rate that looks automated. A proxy changes the exit. The request still looks automated.

What a proxy does not authorize

Both route through a real person's connection. That does not give you permission to access a site, bypass authentication or a paywall, or ignore robots rules and terms of service, and it does not make unlawful collection lawful. If the plan only works because the target believes you are someone else, a proxy will not fix it.

Frequently Asked Questions

What is the difference between a mobile proxy and a residential proxy?

A residential proxy exits through an ISP address assigned to a home connection. A mobile proxy exits through a carrier gateway, usually behind CGNAT, where many phones share one public IP. The trust, price, pool size, and latency differences all follow from that.

Which is better for web scraping?

Residential is the better default for bulk scraping. It costs less per GB and the pool is larger. Mobile is for the subset of requests that fail on residential ranges, because carrier ASNs are expensive for a target to block without collateral damage.

Why are mobile proxies more expensive?

Carrier bandwidth costs more, the pool of carrier IPs is smaller, and the carrier ASN carries the highest trust score. ProxyBase mobile starts at about $5 per GB against about $3 per GB for residential. Rates move by country and supply.

Are mobile proxies anonymous?

No. Both types are real consumer connections, and mobile IPs are shared through carrier-grade NAT, so reputation is shared too. A proxy changes the exit address. It does not change the headers, cookies, account, or behavior the target can see.

Should I use a sticky or rotating mobile session?

Sticky for logins, carts, and multi-step flows. Rotating for independent URLs where no state carries across requests. Rotating sessions choose a new seller per SOCKS5 connection, so keep-alive can pin one exit.

How do I switch between residential and mobile on ProxyBase?

Open a session with POST /v2/sessions and set network_type to residential or mobile, with the country in the body. Connect to api.proxybase.xyz:1082 with the session id as the username and the session token as the password.

Related reading

Browse Residential and Mobile ProxiesRead the Python Setup Guide