# IPv6 Proxies for Web Scraping: What Works

> IPv6 proxies for web scraping from $0.083/GB or unmetered from $10 a day: 48 of 127 popular targets reachable over IPv6, setup, pacing per /64, costs.

[Home](https://quanticdata.io/) / [IPv6 Proxies](https://quanticdata.io/ipv6-proxies/) / For web scraping

IPV6 PROXIES · WEB SCRAPING

# IPv6 proxies for web scraping

**IPv6 proxies are the cheapest way to scrape sites that answer over IPv6; for the rest, our IPv4 proxies run on the same account.** In our DNS census of 18 November 2026, 48 of 127 popular targets were reachable over IPv6, including Google, Bing, YouTube, Amazon's www host and Expedia. IPv6 bandwidth costs $0.20 to $0.083 per GB, or nothing per GB on a speed plan from $10 a day.

**From $0.083**/GB IPv6 3 TB · $250 total
or unmetered from $10 for 1 day

[Buy IPv6 proxies](https://app.quanticdata.io/register?next=%2Fplans%3Fnetwork%3DIPV6&utm_source=wave6-ipv6&utm_medium=ai-twin&utm_campaign=ipv6&utm_content=ipv6-scraping-hero) · [See the targets](https://quanticdata.io/ipv6-proxies-for-web-scraping/#targets)

Mixed list? IPv4-only targets go through datacenter from $0.50/GB on the same account.

- Visa
- Mastercard
- American Express
- Apple Pay
- PayPal
- Bitcoin · Ethereum · Tether (USDT) · 300+ cryptocurrencies

Pay as you go · No subscription · No KYC

**26** countries

[US IPv6 proxies](https://quanticdata.io/ipv6-proxies/united-states/)

Per-GB packs

From $0.20/GB at entry down to $0.083/GB at 3 TB.

Unmetered from $10 for 1 day

The limit is the port speed you choose, 30, 60, 120 or 200 Mbps, for 1, 7 or 30 days.

rotating · static

Rotating & static sessions

A rotating session takes a new IPv6 address on every request; a static session keeps one address for the session.

HTTP · HTTPS · SOCKS5

HTTP, HTTPS and SOCKS5, with a username and password, on packs and speed plans alike.

By [Aldo Morese](https://quanticdata.io/about/), founder of QuanticData · Published Nov 18, 2026 · Last updated Nov 18, 2026 · Target data: [DNS census of 127 sites](https://quanticdata.io/websites-that-support-ipv6-proxies/)

- [Which targets you can scrape](https://quanticdata.io/ipv6-proxies-for-web-scraping/#targets)
- [Setup in three steps](https://quanticdata.io/ipv6-proxies-for-web-scraping/#setup)
- [Rotation and pacing](https://quanticdata.io/ipv6-proxies-for-web-scraping/#pacing)
- [What it costs](https://quanticdata.io/ipv6-proxies-for-web-scraping/#cost)
- [When to use IPv4](https://quanticdata.io/ipv6-proxies-for-web-scraping/#when-not)
- [FAQ](https://quanticdata.io/ipv6-proxies-for-web-scraping/#faq)

From our census of 18 November 2026

## Which scraping targets an IPv6 proxy reaches

Before anything else, split your target list in two. A site whose landing host has an IPv6 address can be scraped through an IPv6 exit; a site without one goes to an IPv4 endpoint. These are the sites of our census that are reachable, by category; the [full list of 127](https://quanticdata.io/websites-that-support-ipv6-proxies/) has the IPv4-only ones too.

Sites reachable over IPv6 on the landing host, DNS census of 18 November 2026 ([CSV](https://quanticdata.io/websites-that-support-ipv6-proxies/ipv6-site-census-2026-11-18.csv))

| Category | Reachable | Scrape over IPv6 |
| --- | --- | --- |
| Search engines | 9 of 14 | google.com, bing.com, yahoo.com, yandex.com, ecosia.org, search.brave.com, seznam.cz, ask.com, aol.com |
| Social networks and video | 12 of 28 | facebook.com, instagram.com, youtube.com, linkedin.com, threads.com, weibo.com, telegram.org, whatsapp.com, mastodon.social, douyin.com, kick.com, line.me |
| Marketplaces and retail | 7 of 31 | amazon.com, lazada.sg, taobao.com, otto.de, wish.com, vinted.com, stockx.com |
| Travel | 7 of 19 | expedia.com, hotels.com, kayak.com, trip.com, trivago.com, ryanair.com, lufthansa.com |
| Classifieds, property, autos, jobs | 6 of 21 | kleinanzeigen.de, olx.pl, autotrader.com, autotrader.co.uk, cars.com, mobile.de |
| Ad platforms | 7 of 14 | ads.google.com, ads.microsoft.com, business.linkedin.com, ads.reddit.com, criteo.com, adsense.google.com, facebook.com/business/ads |

Reachable means the host publishes an AAAA record. It does not mean the site serves scrapers without limits: test your own pages before moving volume.

## How to scrape with IPv6 proxies, in three steps

1. **Check your targets.** Run the open-source [ipv6-reachability](https://github.com/quanticdata-io/ipv6-reachability) checker on your list. It follows redirects and checks every host, so it catches the cases where the bare domain and the www host differ, such as Amazon.
2. **Request the final URL.** For a site marked "yes, from www" in our list, point the scraper at the www URL. A bare domain with no IPv6 address would not resolve over IPv6.
3. **Route the list by protocol.** Send reachable targets through the IPv6 endpoint and the rest through a [datacenter](https://quanticdata.io/datacenter-proxies/) or [residential](https://quanticdata.io/residential-proxies/) endpoint on the same account.

```
# 1. which targets answer over IPv6
ipv6-reachability sites --doh https://cloudflare-dns.com/dns-query -f targets.txt

# 2. scrape a reachable one through the IPv6 endpoint
curl -x ipv6.quanticdata.io:1001 \
  -U "USER-country-us-package-ipv6:PASS" \
  https://www.amazon.com/
```

Endpoint and username format as documented on the [IPv6 proxies page](https://quanticdata.io/ipv6-proxies/#choose-setup); the country code goes in the username (26 countries plus a worldwide mix).

## Rotation and pacing: budget per /64, not per address

A rotating IPv6 session gives every request a new address. That helps against limits counted per address, but large sites usually count IPv6 traffic per `/64` prefix, the block one customer normally gets. Plan your request rate per /64, spread the job over time, and keep the request rate of each prefix close to what a person would do.

Our exits are drawn from two /29 IPv6 ranges, over a million /48 subnets, not from one subnet assigned to you, so consecutive requests rarely share a /64. Static sessions keep one address for flows that need it, such as paging through results with a cookie. Up to 10,000 connections can run at once on every IPv6 plan.

## What scraping over IPv6 costs

Two ways to pay. Per GB, the 100 GB pack is $20 ($0.20/GB) and the 3 TB pack is $250 ($0.083/GB), each valid 30 days, with no subscription: an idle day costs nothing. By port speed, a speed plan has no traffic meter at all for its term: from $10 for one day at 30 Mbps to $350 for 30 days at 200 Mbps.

Take a pack when the job has a known size and runs in bursts; take a speed plan when a crawler will keep the port busy most of the day. Either way, the same gigabyte on datacenter IPv4 costs $0.50 and on residential from $0.80: our IPv6 page compares 1 TB at $110 on IPv6 with about $500 on datacenter.

IPv6 speed plans, total per term, traffic not metered (grid read 2026-11-16)

| Term | 30 Mbps | 200 Mbps |
| --- | --- | --- |
| 1 day | $10 | $25 |
| 7 days | $45 | $120 |
| 30 days | $110 | $350 |

60 and 120 Mbps and every other cell: [the full grid](https://quanticdata.io/ipv6-proxies/#speed-plans).

## When to scrape with our IPv4 proxies instead

- **The target is IPv4-only.** 79 of the 127 sites in our census, among them eBay, Walmart, AliExpress, Booking.com, Airbnb, Zillow, Indeed, TikTok and X. Use [datacenter](https://quanticdata.io/datacenter-proxies/) first, then residential if the site asks for a household address.
- **The site asks for a household address.** Our IPv6 exits are datacenter-issued; a site that wants home addresses needs [residential proxies](https://quanticdata.io/residential-proxies/).
- **You want results, not exits.** Our [SERP API](https://quanticdata.io/serp-api/) returns Google and Bing results as JSON and the [Web Scraping API](https://quanticdata.io/web-scraping-api/) returns any page as Markdown or HTML, with the proxies handled on our side.

## FAQ

### Are IPv6 proxies good for web scraping?

Yes, on targets that answer over IPv6: they are the cheapest bandwidth we sell, $0.20 to $0.083 per GB, with a new address per request. On 18 November 2026, 48 of the 127 popular sites we checked were reachable over IPv6, including Google, Bing, YouTube and Amazon's www host. For IPv4-only sites, such as eBay, Walmart or Booking.com, use our datacenter or residential proxies.

### Can I scrape Google with IPv6 proxies?

Yes. www.google.com and the autocomplete host suggestqueries.google.com both publish IPv6 addresses (checked 18 November 2026), and so do Bing and Yahoo. Google rate-limits IPv6 traffic per /64 prefix, so pace requests per /64 rather than per address. If you would rather receive the results as JSON, our SERP API does the fetching.

### Can I scrape Amazon with IPv6 proxies?

Yes, from the www host. amazon.com has no IPv6 address, but www.amazon.com, www.amazon.de and the image host m.media-amazon.com do (checked 18 November 2026). Request https://www.amazon.com/ directly so the first hop does not go to the bare domain.

### Rotating or static IPv6 proxies for scraping?

Rotating for most scraping: each request leaves from a new IPv6 address, drawn across many /48 and /64 subnets. Static when a flow needs the same address for its whole session, such as paging with a cookie or a cart. Both run on the same IPv6 plans, over HTTP, HTTPS and SOCKS5.

### How many IPv6 proxies do I need for scraping?

You do not buy a number of addresses: you buy traffic (from 100 GB for $20) or port speed (from $10 a day), and every request can take a new address. What to size is concurrency and pacing: up to 10,000 connections at once, and a request rate per /64 the target tolerates.

## Scrape the IPv6 half of your list for less

IPv6 from $20 for 100 GB down to $0.083/GB, or a speed plan with traffic not metered from $10 for 1 day: rotating or static, 26 countries, up to 10,000 connections. The IPv4-only half runs on [datacenter](https://quanticdata.io/datacenter-proxies/) or [residential](https://quanticdata.io/residential-proxies/) exits on the same account.

[Buy IPv6 from $20](https://app.quanticdata.io/register?next=%2Fplans%3Fnetwork%3DIPV6&utm_source=wave6-ipv6&utm_medium=ai-twin&utm_campaign=ipv6&utm_content=ipv6-scraping-closing) · [See every IPv6 plan](https://quanticdata.io/ipv6-proxies/#plans)

## Related pages

- [Which sites support IPv6 proxies 127 sites checked](https://quanticdata.io/websites-that-support-ipv6-proxies/)
- [IPv4 vs IPv6 proxies Price, pool and sites](https://quanticdata.io/ipv4-vs-ipv6-proxies/)
- [IPv6 proxies From $0.083/GB](https://quanticdata.io/ipv6-proxies/)
- [IPv6 proxies for SEO Engines that accept them](https://quanticdata.io/blog/ipv6-proxies-for-seo/)
- [IPv6 proxies for Google Jobs to run on an IPv6 exit](https://quanticdata.io/ipv6-proxies/for/google/)
- [Web Scraping API Pages as Markdown, proxies handled](https://quanticdata.io/web-scraping-api/)

---

Source: https://quanticdata.io/ipv6-proxies-for-web-scraping/ · Site index for AI: https://quanticdata.io/llms.txt
