ASN-Targeted Proxies: When You Need a Specific ISP Network

The internet routes traffic between autonomous systems, each identified by an ASN and each announcing the IP prefixes it is responsible for through BGP. An ASN filter on a proxy product says: only give me exits whose IP falls inside prefixes originated by this AS.

That is a precise statement about routing. It says nothing directly about:

• Location within the network. A national ISP's ASN covers the whole country. Add a city filter if location matters, and expect the pool to shrink sharply.
• Customer type. The same ASN can carry home broadband, business lines and hosted services the ISP runs itself.
• How the IP got into the proxy pool. A residential exit on the right ASN may be a consenting app user's home connection, or, in the ISP proxy category, an address the ISP leased out and that is actually hosted in a datacenter. Our explainer on BGP and RIR IP allocations covers how that happens.

Keep that distinction in mind. A site can see the ASN of your exit. It can also see things an ASN filter does not control, such as latency patterns, the organisation name on the address record, and whether the IP has a history.

The ISP-blocking row is the one teams underestimate. Brand protection and anti-piracy teams need to know whether a blocking order is actually applied by each named ISP, and the only honest test is to request the blocked domain from an exit on each network. A country filter mixes all ISPs together and averages the answer away.

Scraping public pages. Anti-bot systems care whether an ASN is residential, mobile or hosting, and how an IP has behaved. They rarely care which residential ISP it is. Adding a Comcast filter to a US residential pool does not make it more trusted than a Charter exit. It makes the pool smaller.

"Premium" feeling. An ASN filter narrows the set of IPs you rotate through, so the same addresses come back sooner and per-IP rate limits bite earlier. On per-gigabyte plans you also pay for the extra retries.

Avoiding datacenter detection. If hosting ASNs are being blocked, the fix is a different proxy type, not a specific ISP. Moving to residential or ISP exits at country level solves it. ISP proxies vs residential proxies helps pick between those two.

Public pages that do not block hosting ranges at all. Many targets are fine with datacenter traffic. There, a flat datacenter plan is the cheapest option by a wide margin, and paying for residential ASN filters is solving a problem you do not have.

IP intelligence services and proxy dashboards show several similar-looking fields, and vendors do not always filter on the one they name.

Three practical consequences:

1. A filter named "ISP: Comcast" may be matching an intelligence vendor's ISP field, not the ASN list. Ask which.
2. Large ISPs often run more than one ASN, for regional networks, acquisitions or business services. A filter on a single ASN can miss consumer ranges that sit elsewhere, and a filter on the brand can include business or hosting ranges you did not want.
3. A target that checks organisation rather than ASN can see a business name on an IP that is technically inside a consumer ISP's ASN. If your task depends on looking like a home user, check the organisation and usage-type fields on your exits too.

One more complication: the mapping changes over time. IPv4 addresses are bought, sold and transferred between organisations, and a prefix can start being announced by a different ASN overnight. BGP reflects the move immediately. IP intelligence databases catch up on their own update schedules, sometimes weeks later. So for a while the same IP can show the new ASN in routing data and the old ISP name in a lookup service, and a vendor's filter can follow either one. If your results depend on a specific network, prefer routing data as the source of truth, and for a surprising exit check RIPEstat's routing history for that prefix to see whether it recently changed hands.

ASN targeting turns up in three product shapes.

Rotating residential pools with an ASN parameter

The most common form. You add an ASN (or ISP) value to the proxy username or pick it in a dashboard, and each request or session exits from a residential device on that network. Bright Data's mobile proxy page describes its proxy services as able to target "any country, city, ZIP Code, carrier, & ASN", and SOAX's mobile proxy page advertises "Target by region, city, or ASN" (both as of September 2026). Confirm on each vendor's residential or ISP product page that the same filter applies to the product you are buying. Pricing is usually per gigabyte, and the ASN filter itself may be included, limited to certain plans, or priced higher. Check the plan comparison on the vendor's page.

Weak point: pool depth for smaller ISPs. A regional provider may have only a small number of available devices at any time, so sessions drop or fall back.

Static ISP proxies sold by provider

ISP proxies are static addresses registered to a consumer or business ISP but usually hosted on servers. Vendors often let you choose the provider when ordering, which is effectively ASN targeting at purchase time. The IP stays fixed, which suits account work where the network must not change. What are ISP proxies explains the category.

Weak point: choice is limited to whichever ISPs the vendor has ranges from, typically a handful per country, and the ranges can carry the history of other customers.

ASN allowlists on enterprise contracts

Large buyers sometimes negotiate a fixed list of ASNs for a project, for example a set of ISPs in one market for a compliance audit. This behaves like a filtered pool with a contract around availability.

Weak point: minimum commitments, and the list has to be maintained as networks merge or renumber.

Some buyers searching for ASN targeting actually want the reverse: to avoid concentration in one network.

If a target rate-limits or blocks a whole ASN, which happens with hosting providers, every exit you have in that ASN fails together. A pool whose exits sit in many ASNs degrades gradually instead of all at once. This matters most for datacenter proxies, where exits are by definition in hosting networks and big cloud ASNs such as Amazon's AS16509 are well known to anti-bot vendors. Cloud provider IP ranges explains why those ranges are treated differently from smaller hosting networks.

Vendors rarely sell "exclude ASN X" as a feature, so diversity is something you measure rather than select. Sample a few hundred exits during a trial and count distinct ASNs and distinct /24 prefixes. Our guide to clean datacenter subnets covers the subnet side of the same idea.

This is also where SparkProxy fits, and where it does not. Our plans give rotating access to 1M+ datacenter IPs across 80+ countries, with random per-request rotation or 5-minute auto rotation, starting at $75 a month for 100 threads with unlimited bandwidth. We do not sell a filter for a specific ISP or ASN, and datacenter exits will not pass as consumer broadband. If your job needs a named ISP, buy residential or ISP proxies from a specialist. If it needs volume on targets that accept datacenter traffic, measure our exits' spread during a trial the same way you would any vendor's.

Whatever the vendor promises, log the actual ASN of your exits. RIPEstat, run by the RIPE NCC, has a free public data API that returns the originating ASN and prefix for any IP, globally, not just in Europe.

Check a single exit:

# 1. Find the exit IP through your proxy
IP=$(curl -s -x http://USERNAME:PASSWORD@PROXY_HOST:PORT https://api.ipify.org)

# 2. Which AS originates the prefix containing it?
curl -s "https://stat.ripe.net/data/network-info/data.json?resource=$IP" | jq '.data'

# 3. Who holds that AS? (replace 7922 with the ASN returned above)
curl -s "https://stat.ripe.net/data/as-overview/data.json?resource=AS7922" | jq '.data.holder'

Then sample the filter over time. This loop records the exit IP, prefix and origin ASN for 100 requests into a CSV you can pivot:

echo "ip,prefix,asn" > exits.csv
for i in $(seq 1 100); do
  IP=$(curl -s --max-time 30 -x http://USERNAME:PASSWORD@PROXY_HOST:PORT https://api.ipify.org)
  [ -z "$IP" ] && continue
  curl -s "https://stat.ripe.net/data/network-info/data.json?resource=$IP" \
    | jq -r --arg ip "$IP" '[$ip, .data.prefix, (.data.asns | join("|"))] | @csv' >> exits.csv
  sleep 2
done

# Summary: requests per origin ASN, and distinct IPs and prefixes seen
tail -n +2 exits.csv | cut -d, -f3 | sort | uniq -c | sort -rn
echo "distinct IPs:      $(tail -n +2 exits.csv | cut -d, -f1 | sort -u | wc -l)"
echo "distinct prefixes: $(tail -n +2 exits.csv | cut -d, -f2 | sort -u | wc -l)"

How to read it:

• Every row shows the ASN you asked for: the filter is honest. Next look at distinct IPs; a low number means a thin pool.
• Some rows show a different ISP: the vendor falls back when the requested network is unavailable. Ask if that can be turned off, because it silently corrupts ISP-specific tests.
• Rows show a hosting ASN on a residential product: raise it with the vendor immediately.

RIPEstat tells you routing truth. For how IP intelligence services label the same IP (ISP name, organisation, usage type), cross-check with one or two commercial lookup services. The guide on validating IP geolocation and fraud scores covers that side.

1. Is the filter keyed on ASN, on an ISP name from an IP database, or on registry organisation?
2. For a multi-ASN ISP, which ASNs are included? Get the list.
3. How many exits are available on this ASN right now, in the country and city I need?
4. What happens when none are available: error, wait, or fallback to another network?
5. Is ASN targeting included in my plan, or priced separately?
6. For static ISP proxies, which registry organisation appears on the addresses, and have they been used by other customers?
7. Can I run a trial with the filter enabled, long enough to sample exits at the target's peak hour?

Pricing for residential traffic varies widely by vendor and commitment, and residential proxy pricing explained covers how per-gigabyte plans compare.