Use cases
Proxies for Web Scraping
Most scraping jobs fail on addresses long before they fail on parsing. Too few origins and too little diversity give the target a pattern it has already learned to recognise.
What breaks without the right network
A scraper running from one range shows the target the same origin on every request, which is the signal a rate limiter or a bot filter is built to catch.
- Sites that serve different prices, catalogues, or availability by geography
- Per-IP rate limits that throttle a single origin long before the job finishes
- Outright blocks on ranges the target has already flagged as automated
- A sample skewed toward whatever one address happened to see
What changes the outcome
Enough addresses that rotation means something
The residential pool runs to 38M+ addresses across 190+ countries, so a rotating job is drawing from a pool wide enough that no single target sees the same origin twice in a session.
Sessions that hold when a flow needs them to
Sticky sessions to 60 minutes keep a multi-step crawl (pagination, a login, an add-to-cart flow) on one address for as long as the flow needs, then release it back to rotation.
A cheaper lane for volume that does not need to blend in
Datacenter addresses are dedicated to one customer at a time and unmetered at 100 Mbps per address. We may step in if usage interferes with other customers. Use them once volume matters more to a job than how the target blocks.
One gateway and one credential format for both
Residential and datacenter sit behind the same HTTP, HTTPS, and SOCKS5 gateway, so moving a job between them means changing one parameter.
Coming soon
A request-metered Scraper API (one call in, HTML back, no gateway to configure) is coming soon for teams that prefer not to operate a pool themselves. It is not yet priced or available.
Which network to put this on
Primary
Residential
Broad, rotating address diversity is what keeps a scraping job under a rate limiter's radar and shows each target a plausible, geographically consistent visitor.
from $2.40 / GB
Alternative
Datacenter
For high-volume collection where the target does not discriminate by origin type, dedicated datacenter throughput is the cheaper way to move the same number of pages.
from $1.50 / IP / month
A workflow this fits into
- 01Point the collection job at the gateway instead of a fixed list of addresses
- 02Rotate per request for broad coverage, or hold a sticky session for a multi-step flow
- 03Watch success and block rates per target, and move a target between residential and datacenter based on what it responds to
- 04Hand the returned HTML to parsing and storage downstream, unchanged
Publicly available data, used lawfully
Every recommendation above assumes lawful collection of publicly available data. Our acceptable use policy governs what is and is not permitted on this network, and sourcing and compliance explains where the addresses themselves come from.
Questions we get about this
Will scraping through this network get me blocked?
No network eliminates blocking entirely. Most block rules are designed to catch traffic with little address diversity and no session control, and residential rotation and datacenter throughput both reduce the odds materially compared with running from a handful of static addresses.
Do you rotate IPs automatically, or do I control it?
Both. The default is rotation per request; sessions can be held sticky for up to 60 minutes on residential when a flow needs continuity. You choose per request.
Can I choose which country a request appears to come from?
ISP and datacenter support country-level targeting. Residential and mobile provide broad international coverage without a per-request country selector at checkout.
Is collecting public web data through a proxy lawful?
Collecting publicly available data is generally lawful, but the target site's own terms and applicable law still apply, and the use has to be one our acceptable use policy permits. Read the acceptable use policy before you scope a job.
What is the difference between residential and datacenter here?
Residential addresses come from consented peers and carry consumer-ISP reputation, which is what a target's blocking logic is usually tuned around. Datacenter addresses are dedicated and much faster, at the cost of that reputation. That trade makes sense when the target does not filter by origin type.
Other use cases
Price Monitoring
See the price, stock, and range a shopper in each market sees.
View use case
SEO and SERP Tracking
Measure rankings the way a searcher in that market and on that device sees them.
View use case
Ad Verification
Confirm creatives render where they were bought, from a real consumer vantage point.
View use case
Ready to run web scraping on the right network?
Create an account and order from 1 GB or 1 IP, or talk to sales if you want the volume ladder priced against your own numbers first.
One business day, from a named engineer.