# Proxies for Web Scraping – Rotating Residential from $0.80/GB

Proxies for web scraping: rotating residential from $0.80/GB, datacenter from $0.50/GB, ISP IPs for logged-in sessions. 200+ countries, pay as you go.

## Proxies for Web Scraping: Rotating Residential, Datacenter and ISP IPs

**Rotating residential from $0.80/GB · datacenter from $0.50/GB · 200+ countries**

The right proxies for web scraping depend on the target: datacenter IPs from $0.50/GB for open sites, rotating residential IPs from $0.80/GB for retailers, marketplaces and other protected pages, and ISP proxies from $3.80 per IP per month for logged-in sessions. Quantum Proxies sells all three pay as you go, over HTTP or SOCKS5, with country and city targeting in 200+ countries.

- **90M+** residential IPs in the rotating pool
- **$0.80** per GB Residential Basic at 1,000 GB
- **$0.50** per GB datacenter at 500 GB
- **100 MB** free trial, no card, unlocked by SMS

## Which proxy type for which scraping target

Start from the site, not the proxy. Public pages that do not fight automation, such as government portals, documentation, sitemaps, RSS feeds and many B2B catalogues, scrape well from datacenter IPs: 1 Gbps+ lines, rotating or dedicated addresses, and $0.80 per GB on the 50 GB plan falling to $0.50 per GB at 500 GB. Retailers, marketplaces, travel sites, classifieds and search engines score datacenter ranges as automation, so those targets want rotating residential IPs from real household connections.

The third case is a scraper that logs in and stays logged in: a dashboard, an account export, a paginated report behind authentication. There you want the same address every time, which is what an ISP proxy gives you, a static residential IP hosted on a datacenter line with unmetered bandwidth for $3.80 per IP per month. Mobile 4G/5G IPs cover the narrow set of targets that serve different content to carrier traffic.

- Datacenter: open sites, APIs, feeds, documentation, from $0.50/GB
- Rotating residential: retailers, marketplaces, travel, classifieds, search, from $0.80/GB
- ISP / static residential: logged-in sessions and dashboards, $3.80 per IP per month
- Mobile 4G/5G: content served differently to carrier traffic

## Rotating vs sticky sessions inside a scraper

A rotating session gives every request a new IP from the pool. It is the right default for crawling many independent URLs: product listings, search results, article archives. No single address accumulates a request count that looks unusual, and a blocked IP costs you one retry rather than the whole run. On Residential Basic and Premium, rotation per request is the default behaviour of the gateway.

A sticky session keeps one IP for a defined time so the site sees a single visitor. Use it when the scraper walks a multi-step flow: log in and then read pages, add to cart to reveal a price, step through numbered pages that share a server-side session. Residential Basic holds a sticky IP from 3 minutes, Residential Premium up to 120 minutes and mobile up to 24 hours. Both modes are selected per request in the proxy username, so one scraper can mix them.

## Concurrency and geo targeting

Scrapers are usually limited by how many requests they can run in parallel, not by bandwidth. Residential Premium allows unlimited concurrent connections and datacenter lines run at 1 Gbps+, so the practical ceiling is the politeness you choose per target site. A common pattern is a fixed number of workers per domain, each on a rotating session, with a short delay between requests to the same host.

Geo targeting matters when the page changes with the visitor's location: prices in local currency, shipping availability, region-specific search results, content that only serves one country. Residential Basic covers 200+ countries with city and ASN targeting; Residential Premium adds state and ISP selection; datacenter proxies target country, state and city across 100+ countries. The location is a parameter in the username, so a single scraper can request the German version of a page and the US version back to back.

## What 1,000 pages cost

The arithmetic is simple once you fix a page size. A typical HTML page without images is roughly 60 KB, so 1 GB ≈ 1,000,000 KB ÷ 60 ≈ 16,000 pages; call it roughly 15,000 to allow for heavier templates. On Residential Basic at $0.80–$1.00 per GB, 1,000 pages therefore cost roughly $0.05–$0.07. On datacenter at $0.50–$0.80 per GB, roughly $0.03–$0.05. On Residential Premium at $2.20–$2.50 per GB, roughly $0.14–$0.16.

Two habits keep you at the low end. Request HTML only and skip images, fonts and tracking scripts, which a headless browser will otherwise download through the proxy. And decide which targets really need residential IPs; moving open sites to datacenter cuts the bill for that share of traffic. Bandwidth is bought once, stays valid for 90–360 days, and there is no subscription.

## Plugging the proxies into Python, Scrapy, Playwright and curl

Every product is a standard HTTP/HTTPS or SOCKS5 proxy with user:pass authentication or IP whitelisting, so integration is a one-line change in the tool you already use. In Python requests or httpx you pass the gateway URL with credentials in the proxies argument; Scrapy reads the same URL from the request meta or a downloader middleware, which is also where you set per-spider rotation. Playwright and Puppeteer take the server and credentials as launch options, and every page in that browser routes through them.

From the shell, curl accepts the proxy with the -x flag plus -U for the credentials, which is the fastest way to confirm a target renders before writing code. Session type, country and city are encoded in the username, so switching from rotating to sticky or from a US to a UK exit is a string change, not a code change.

## SERP API and Extract API: scraping without managing proxies

If your goal is the data rather than the proxy plumbing, two hosted endpoints skip most of this page. The SERP API at /serp-api takes a query and a location and returns Google results as JSON; you never render a results page or handle a captcha yourself. The Extract API at /extract-api fetches any URL through the residential network and returns clean Markdown or JSON, with navigation, scripts and boilerplate removed, which is the format most pipelines and LLMs want anyway.

Both bill from the same wallet as the raw proxies, so a project can start with the APIs for the hard targets and move volume crawling to plain rotating residential or datacenter traffic where that is cheaper. For teams building on Claude or Cursor, the MCP server at /mcp-server exposes the same search, scrape and crawl capabilities as tools.

## Scraping responsibly

Proxies solve the IP problem; they do not replace good scraping manners. Collect data that is public, read the target's terms and robots directives, rate-limit per host, cache what you have already fetched and identify your crawler where the site asks for it. Rotating IPs are there to distribute load and to give you localized views, not to hammer a server that has asked you to slow down.

The pay-as-you-go model fits this: top up the wallet from $5, give each project its own proxy credential, watch consumption per credential in the dashboard, and stop when the job is done. There is a 100 MB free trial on Residential Basic, valid 5 days and unlocked by a one-time SMS verification, so a new target can be tested before buying bandwidth.

## Proxy type vs scraping target

| Target | Proxy type | Session | Price from |
| --- | --- | --- | --- |
| Open sites, public APIs, feeds, documentation | Datacenter | Rotating or dedicated IP | $0.50/GB at 500 GB ($0.80/GB at 50 GB) |
| Retailers, marketplaces, classifieds, travel | Residential Basic | Rotating, country or city targeted | $0.80/GB at 1,000 GB ($1.00/GB at 25 GB) |
| Search engines | Residential Basic or SERP API | Rotating, geo-targeted | $0.80/GB; SERP API billed from the same wallet |
| Multi-step flows: login, cart, paginated sessions | Residential Basic or Premium | Sticky, from 3 min (Basic) or up to 120 min (Premium) | $0.80/GB Basic, $2.20/GB Premium at 500 GB |
| Logged-in dashboards, account exports, one fixed identity | ISP / static residential | Fixed IP, unmetered | $3.80 per IP/month, $2.50 above 100 IPs |
| Content served differently to carrier traffic | Mobile 4G/5G | Rotating or sticky up to 24 h | $2.30/GB at 500 GB ($5 for 1 GB) |
| Pages fed into an LLM or a pipeline | Extract API | Handled for you | Billed from the same wallet as the proxies |

## Frequently Asked Questions

### What are the best proxies for web scraping?

It depends on the target. Datacenter proxies from $0.50 per GB are the best value for open sites that do not block automation. Rotating residential proxies from $0.80 per GB are the practical default for retailers, marketplaces, travel sites and search engines. ISP proxies at $3.80 per IP per month suit scrapers that stay logged in. Most projects end up mixing two of them.

### Should a scraper use rotating or sticky proxies?

Rotate when the URLs are independent, such as product listings or search results, so each request leaves from a different IP and a blocked address costs one retry. Use a sticky session when several requests must belong to one visitor: login then dashboard, add to cart then checkout price, numbered pages sharing a session. Both are chosen per request in the proxy username.

### How much do proxies for web scraping cost per 1,000 pages?

Roughly $0.05–$0.07 on Residential Basic. A typical HTML page is about 60 KB, so 1 GB ≈ 1,000,000 KB ÷ 60 ≈ 16,000 pages, call it roughly 15,000. At $0.80–$1.00 per GB that is a few cents per 1,000 pages; datacenter at $0.50–$0.80 per GB is roughly $0.03–$0.05. Fetching HTML only keeps you at the low end.

### How do I use these proxies with Python requests or Scrapy?

Pass the gateway URL with your username and password as the proxy setting: the proxies argument in requests or httpx, the request meta or a downloader middleware in Scrapy, the launch options in Playwright or Puppeteer. Country, city and session type are part of the username, so a scraper changes location or switches to sticky without any code change.

### VPN or proxy for web scraping?

A VPN routes your whole machine through one IP, which is the opposite of what a scraper needs: every request from the same address, no per-request control, no geo choice per call. A proxy is set per request or per session, rotates across a pool of 90M+ IPs, and can target a country or city in the username. For scraping, use proxies.

### Do I need residential proxies, or is datacenter enough?

Test the target from datacenter first: if it returns the page reliably, keep it there at $0.50–$0.80 per GB. Sites that score datacenter ranges as automation, which includes most retailers, marketplaces, classifieds and search engines, need rotating residential IPs from $0.80 per GB. Splitting your URL list between the two is the cheapest reliable setup.

### Is there a free trial for scraping proxies?

Yes. New accounts get 100 MB of Residential Basic traffic free for 5 days, one per person, unlocked by a one-time SMS verification in Settings → Free Trial, with no card. At roughly 60 KB per HTML page that is enough for a few hundred to over a thousand fetches, which is plenty to confirm a target and a session type before buying.

## Related

- [Residential Proxies](https://quantumproxies.io/residential-proxies)
- [Datacenter Proxies](https://quantumproxies.io/datacenter-proxies)
- [ISP Proxies](https://quantumproxies.io/isp-proxies)
- [Rotating Proxies](https://quantumproxies.io/rotating-proxies)
- [Extract API](https://quantumproxies.io/extract-api)
- [SERP API](https://quantumproxies.io/serp-api)