What PacketStream is: a peer bandwidth marketplace
PacketStream is a two-sided residential proxy marketplace and one of the earliest consumer bandwidth-sharing networks to reach scale. On one side, people install a desktop client and are paid per gigabyte routed through their home connection. On the other, customers buy that bandwidth back at a flat per-gigabyte price. Almost every characteristic of the product — good and bad — follows from that design rather than from a pricing decision.
The product is deliberately narrow. Residential IPs only, one gateway hostname and port, the country chosen by appending a code to the proxy username, and a credit balance you top up and draw down. There is no datacentre line, no static ISP line, no mobile line, and no managed scraping or unblocking layer sitting in front of the pool. For a large class of jobs that narrowness is the point: there is very little to configure and correspondingly little to get wrong.
- Residential IPs only, sourced from a paid peer bandwidth-sharing application
- A single gateway endpoint; country selected inside the username string
- Pay-as-you-go credit, no subscription, purchased balance does not expire
- HTTP, HTTPS and SOCKS5 with username/password authentication
- A white-label reseller API and the paid bandwidth-sharing side are separate offerings
When to stay with PacketStream
If your work is country-level rotation and nothing beyond it — generic collection, price and availability checks across a handful of markets, light SEO sampling — PacketStream does that with less ceremony than almost anything else in the category. A single flat rate with no tiers means the monthly cost is arithmetic you can do in your head, there is no plan to outgrow, and nothing renews on you.
Two further reasons to stay are commercial rather than technical. If you resell proxies under your own brand through their reseller API, that relationship is not something a standard proxy account replaces. If you also sit on the earning side of the network, no competitor offers the same arrangement. And if you are holding a substantial non-expiring balance, the rational move is to spend it down rather than abandon it — running two accounts in parallel for a few weeks costs nothing.
City targeting and session length vs PacketStream
The constraints PacketStream users tend to run into are consistent: you can specify a country and that is the extent of the geo control, and a sticky session persists only while the peer behind it stays online, with a comparatively short ceiling. Neither is an oversight — both are honest consequences of routing through volunteer consumer devices that come and go.
Nefo is built differently on both axes. Targeting covers 227 countries and territories, plus United States state and city, region and city in selected countries such as the United Kingdom, Canada, Germany and France, and a specific ASN or ISP paired with a country. Session behaviour is declared per credential rather than inferred: rotating gives a new IP on each request, sticky holds the same IP for anything from one minute to 24 hours. If you have been compensating for country-only targeting by pulling a broad sample and discarding most of it, that overhead disappears — and the bandwidth it consumed with it.
One caveat worth checking rather than assuming: PacketStream has signalled finer geo-targeting as planned work. Read their current documentation before treating country-only as a permanent property.
- 227 countries and territories at credential-generation time
- United States down to state and city
- Region and city in selected countries (UK, Canada, Germany, France)
- ASN / ISP targeting, paired with a country
- Sticky sessions from 1 minute to 24 hours, configured per credential
- Rotating sessions with a new IP per request
How prepaid PacketStream and Nefo behave under load
Start with what is the same, because it matters and it is often mis-sold: both are prepaid, and on both sides the balance does not expire. Nefo's flexible plans are non-expiring, so "will I lose unused bandwidth" is not a reason to switch either way.
The difference is the shape of the curve and the cost of entry. PacketStream charges one flat rate regardless of how much you buy, behind a minimum opening purchase — so the smallest way in is a block of credit rather than a trial-sized amount. Nefo starts at 5GB ($0.90/GB) and tiers the per-gigabyte rate down as volume rises, to $0.53/GB at 10TB. Practically: flat pricing is easier to forecast and fine if your consumption is modest and steady; tiered pricing is cheaper once you are buying in bulk, and far cheaper to evaluate before you commit, because a real test costs a few dollars rather than a deposit.
Generating credentials on Nefo is free — only traffic passing through them consumes bandwidth. That makes it practical to hold separate credentials per project, per country and per session type, which is how you keep usage attributable rather than reconciling one shared username after the fact.
How to migrate from PacketStream to Nefo
Mechanically this is a small job. Both services are username/password gateways speaking HTTP, HTTPS and SOCKS5, so there are no client library changes and no new authentication model. The one place code usually needs editing is anywhere a country code has been concatenated into the PacketStream username at runtime: on Nefo, targeting is baked into the credential you generate, so that logic moves out of your code and into configuration.
Treat the evaluation as an experiment on your own targets rather than a comparison of published benchmarks. Success rates on residential pools vary enormously by destination site, time of day and country, and no third-party test predicts your workload.
- Buy the 5GB entry plan — enough to characterise a pool, cheap enough to write off
- Generate two credentials: one rotating, one sticky, in the country you use most
- Point a staging worker at them and leave it running across a full 24-hour cycle
- Compare success rate, latency and retry count against your PacketStream numbers on the same targets
- Move production traffic gradually; keep the PacketStream account live until its balance is spent
- Use the dashboard and CSV export to reconcile consumption during the overlap
PacketStream or Nefo: how to choose
Stay with PacketStream if country-level rotation is genuinely all you need, if flat and predictable pricing matters more to you than a lower unit rate at volume, if you use their reseller API, or if you have credit to burn down. It is a long-running service with a simple, coherent design, and simple and coherent is a real feature.
Move to Nefo if you need geography below country level, if you need a session that holds an IP for hours rather than minutes, if you want to test with 5GB instead of a deposit, or if your volume is high enough that a tiered rate compounds into real money.
There is also one category where the comparison ends rather than tightens. If you need carrier IPs — a 4G/5G port on a real carrier modem with one live rotating mobile IP you can hold or change on demand, rented by duration rather than by gigabyte, with US state and carrier targeting — PacketStream does not sell that product at all. Nefo does, alongside the residential pool and on the same account.
Frequently asked questions
For most users, yes. Both use prepaid credit that does not expire and standard proxy authentication over HTTP, HTTPS and SOCKS5. Nefo adds city and ASN targeting, sticky sessions up to 24 hours, volume-tiered pricing and mobile proxies. Stay with PacketStream if its narrow, flat-rate product already fits.
No. Both services use username and password proxy authentication, so your client library and retry logic stay as they are. You change host, port and credentials. Nefo adds targeting through username tokens such as -country-us and -city-losangeles, so code that builds a PacketStream country username needs a small edit.
No subscription. Nefo starts at 5 GB for $4.50 and bandwidth never expires, so a real test costs little and there is nothing to cancel. Per-GB rates fall as packages grow, to $0.53/GB at 10 TB. See pricing.
Yes. Residential proxies can be targeted by country, state, city and ISP/ASN, from 227 countries and territories, 22,417 cities and 14,252 ISPs. Country-only targeting forces you to over-collect and filter, which burns bandwidth you already paid for.
Sticky sessions run from 1 minute up to 24 hours, chosen per proxy line. On a peer-marketplace network a sticky session is bounded by how long that device stays online. Multi-step flows with server-side state, like logins and carts, benefit from the longer window.
Yes, as a separate line PacketStream does not sell. Mobile proxies are 4G/5G connections on real carrier modems, from $3.90/GB across 110 countries, or a dedicated port by the day from $1.00/day in the US.